| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
In order to avoid adding ranges to the wrong scopes due to
inlining they are numbered consecutively for the whole compilation
unit.
Bug 26234
|
|
|
|
|
|
| |
The previous check was incomplete for integer literals in base 10.
Bug 26119
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Consider:
```
struct s { ... } __attribute((aligned(N)));
struct t { ... }
__attribute((aligned(N))) struct t x;
```
In the first case, the aligned attribute should be attached to struct s, so that further references to struct s are aligned.
In the second case, the aligned attribute should be attached to the variable x, because if we attach it to struct t, it will be ignored and cause a warning.
This commit changes the attachment rule so that it treats both cases right.
Extend regression test for "aligned" attribute accordingly, by testing
aligned attribute applied to a name of struct type.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit adds a check to reject type definitions such as
```
typedef __attribute((section "foo")) int fooint;
```
GCC and Clang also reject this as an error.
Without the check, the behavior is somewhat surprising:
```
fooint x; // placed in section "foo"
fooint * x; // placed in default section, attribute "foo" is ignored
```
Note that the following must be accepted:
```
typedef struct { ... } __attribute((packed)) t;
```
The "packed" attribute is correctly attached to the struct type and should not be checked. This is achieved by using `attribute_of_type_no_expand` to get the attributes of the typedef-ed type, excluding the attributes carried by a struct/union or another typedef.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is a second step towards mimicking GCC/Clang's handling of attributes.
This commit introduces a distinction between
- Object-related attributes, such as "section", which apply to the object (function, variable) being defined;
- Name-related attributes, such as "aligned", which apply to the name (object, struct/union member, struct/union/enum tag) being defined.
In particular, "aligned" is now attached to "struct" and "union" definitions, while it used to be "floated up" before.
The C11 _Alignas modifier is treated like an object-related attribute, so that
```
struct s { ... };
_Alignas(64) struct s x;
```
correctly associates the alignment with "x" and not with "struct s", where it would be ignored because it was not part of the original definition of s.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
During elaboration of type declarators, non-type-related attributes such as "aligned" or "section" are "floated up" so that they apply to the thing being declared. For example, consider:
```
__attribute((aligned(16))) int * p;
```
The attribute is first attached to type `int`, then floated up to type `int *`, so that it finally applies to `p`, giving a 16-aligned pointer to int, and not a naturally-aligned pointer to 16-aligned int.
What happens when the non-type-related attribute comes from a typedef?
```
typedef __attribute((aligned(16))) int i16;
i16 * p;
```
CompCert used to expand the typedef then float up the attribute, resulting in `p` being a 16-aligned pointer to int.
GCC and Clang produce a naturally-aligned pointer, so they do not expand the typedef before floating.
The old CompCert behavior is somewhat surprising, and potentially less useful than the GCC/Clang behavior.
This commit changes the floating up of non-type-related attributes so that typedefs and struct/union/enum definitions are not expanded when determining which attributes to float up. This is a first step towards mimicking the GCC/Clang behavior.
|
|
|
|
|
|
|
|
| |
Instead of relying testing that the size of pointers is 64bit the
size of registers should be tested. Also it should be a fatal
error to reverse a long long on an architecture that does not
support reverse 64bit read/writes.
Bug 24982
|
|
|
|
|
|
| |
Catch the exception from a non constant argument of a packed
attribute and print an error.
Bug 24748
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* bug 24268: avoid assertion after reporting error for invalid call to builtin_debug
* bug 24268, remove duplicated warning tag in lexer messages
* bug 24268, fix spelling in array element designator message
* bug 24268, unify 'consider adding option ...' messages
* bug 24268, add spacing for icbi operands
* bug 24268, uniform use of Ignored_attributes class for identical warnings
* bug 24268, unify message for 'assignment to const type' to error from error/fatal error
* bug 24268, in handcrafted.messages, "a xxx have been recognized" -> "a xxx has been recognized"
|
|
|
|
|
|
| |
Since the following offsetof cannot handle bit-fields we should
stop earlier.
Bug 24480
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Refactor common code of alignas.
Instead of working on attributes the function now works directly
on the type since the check always performed an extraction of
attributes from a type.
Bug 23393
* Attach _Alignas to the name.
Bug 23393
* Attach "aligned" attributes to names
So that __attribute((aligned(N))) remains consistent with _Alignas(N).
gcc and clang apply "aligned" attributes to names, with a special case
for typedefs:
typedef __attribute((aligned(16))) int int_al_16;
int_al_16 * p;
__attribute((aligned(16))) int * q;
For gcc, p is naturally-aligned pointer to 16-aligned int and
q is 16-aligned pointer to naturally-aligned int.
For CompCert with this commit, both p and q are 16-aligned pointers
to naturally-aligned int.
* Resurrect the alignment test involving typedef
The test was removed because it involved an _Alignas in a typedef,
which is no longer supported. However the same effect can be achieved
with an "aligned" attribute, which is still supported in typedef.
|
| |
|
|
|
|
|
|
| |
Instead of performing the check only for parameters of function
definitions also perform it for function declarations.
Bug 23393
|
|
|
|
|
|
|
| |
The new diagnostic triggers if an `_Alignas` or an `aligned` attribute
or a `packed` attribute requests an alignment smaller than the natural alignment.
Bug 23389
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
The C11 standard disallows the usage of _Alignas for:
- Bit-field members of struct or union types
- Typedefs
- Function Defintions
- Parameters of functions
It is still allowed to use the gcc attribute for these constructs.
Bug 23391
|
|
|
|
|
|
| |
The check tests whether the standard _Alignas is contained within
a given attribute list.
Bug 23391
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since commit 5963ac4, "aligned" attributes and _Alignas qualifiers
are represented differently, causing them to be treated differently
by the previous implementation of Cutil.attr_array_applicable.
This is incorrect and inconsistent with what happens during elaboration
of array types in Elab.
This PR reimplements attr_array_applicable in terms of class_of_attribute.
Just like during elaboration, attributes of the Attr_type class
are applied to the type of array elements, other attributes stay
attached to the array type.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We used to recognize attribute(("aligned"(N))) and map it to _Alignas(N)
during elaboration.
However, we want to restrict the places where _Alignas can occur, as
standardized in ISO C11, while leaving more freedom for the placement
of the "aligned" attribute.
As a first step in this direction, this commit keeps the "aligned"
attribute unchanged in the AST, and distinct from _Alignas attributes.
Both attributes are honored when it comes to determining the actual
alignment of a type.
|
|
|
|
|
|
| |
Restrict is only allowed for pointers whose referenced type is an
object type or incomplete type, but not a function type.
Bug 23397
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Add diagnostic for type qualified arrays that occur in the wrong place
Arrays with type qualifiers (e.g. int t[const 5]) are only allowed as
function parameters and for them
only the outermost array type derivation.
Bug 23400
* Keep attributes from array for argument conversion
Type qualifiers of arrays in function parameters are just syntactic sugar
to allow adding them to the resulting pointer type. Hence, when a
qualified array type such as `int t[const 5]` decays into a pointer type
during argument conversion, the pointer type should be qualified, e.g. `int * const t`.
|
|
|
|
|
|
| |
Tentative static definitions with incomplete type are not allowed
in C99. However most popular compilers support them and warn
about them.
Bug 23377
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Error for structs with only flex array member
Flexible array members are only allowed if another member exists.
Bug 23324
* Added checks for nesting of structs with flex array members
Warn if a struct with a flex array member is used as array element
or member of another struct. Such usage is dubious.
Bug 23324
Don't warn if the struct-with-flex-array is a member of an union.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since the size of integer registers is not identical to the size of pointers
for the ppc64 and e5500 model the check for register pairs in
ExtendedAsm does not work correctly.
In order to avoid this a new field sizeof_intreg is introduced in the
Machine configuration which describes the size of integer registers.
New configurations for the ppc64 and e5500 model are added
and used.
Bug 24273
|
|
|
|
|
|
| |
Since the parameter name gets used in other error messages it
results in messages without names.
Bug 24283
|
|
|
|
|
|
|
| |
It's meant as a Boolean (byte-swap or not), so any other value is dangerous.
The error message is the generic "ill-formed 'packed' attribute".
Maybe we don't need a custom error message.
|
|
|
|
|
| |
The list of arguments to the attribute was missing a reverse, hence
attribute(("foo"(1,2,3))) was actually read as attribute(("foo"(3,2,1))).
|
|
|
|
| |
__builtin_offsetof(struct s, f) is an error if f is a bit-field.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
CompCert has two implementations of sizeof, alignof and offsetof (byte offset of a struct field):
- the reference implementation, in Coq, from cfrontend/Ctypes.v
- the implementation used during elaboration, in OCaml, from cparser/Cutil.ml
The reference Coq implementation is used as much as possible, but sometimes during elaboration the size of a type must be computed (e.g. to compute array sizes), or the offset of a field (e.g. to evaluate __builtin_offsetof), in which case the OCaml implementation is used.
This causes issues with packed structs. Currently, the cparser/Cutil.ml functions ignore the "packed" attribute on structs. Their results disagree with the "true" sizes, alignments and offsets computed by the cfrontend/Ctypes.v functions after source-to-source transformation of packed structs as done in cparser/PackedStruct.ml. For example:
```
struct __packed__(1) s { char c; short s; int i; };
assert (__builtin_offsetof(struct s, i) == 3);
assert (sizeof(struct s) = sizeof(char[sizeof(struct s)]));
```
The two assertions fail. In the first assertion, __builtin_offsetof is elaborated to 4, because the packed attribute is ignored during elaboration. In the second assertion, the type `char[sizeof(struct s)]` is elaborated to `char[8]`, again because the packed attribute is ignored during elaboration, while the other `sizeof(struct s)` is computed as 7 after the source-to-source transformation of packed structs.
This commit changes the cparser/Cutil.ml functions so that they take the packed attribute into account when computing sizeof, alignof, offsetof, and struct_layout.
Related changes:
* cparser/Cutil: add `packing_parameters` function to extract packing info from attributes
* cparser/Cutil: refactor and share more code between sizeof_struct, offsetof, and struct_layout
* cparser/Elab: check the alignment parameters given in packed attributes. (The check was previously done in cparser/PackedStruct.ml but now it would come too late.)
* cparser/Elab: refactor the checking of alignment parameters between _Alignas, attribute((aligned)), __packed__, and attribute((packed)).
* cparser/PackedStructs: simplify the code, some functionality was moved to cparser/Cutil, other to cparser/Elab
* cfrontend/C2C: raise an "unsupported" error if a packed struct is defined and -fpacked-structs is not given. Before, the packed attribute would be silently ignored, but now doing so would cause inconsistencies between cfrontend/ and cparser/.
* test/regression/packedstruct1.c: add tests to compare the sizes and the offsets produced by the elaborator with those obtained after elaboration.
|
|
|
|
|
| |
It is not allowed in C to have a parameter in a parameter list
without an identifier.
Bug 24283
|
|
|
|
|
|
| |
Since a non modifiable lvalue is an invalid asm output it should
be checked earlier, otherwise this leads to a retyping error
later.
Bug 24285
|
|
|
|
|
|
| |
Parameters also need to be checkd for unknown attributes, like
all other declarations.
Bug 24277
|
|
|
|
|
|
| |
Fix various typos in diagnostic messages and unified wording and
capitalization.
Bug 23850
|
|
|
|
| |
better name for generated test cases.
|
| |
|
|
|
|
| |
The `_Alignas(expr)` construct is not C11, only `_Alignas(type)` is.
|
| |
|
|
|
|
|
|
|
| |
Nonstatic inline functions can be expanded in several compilation units.
The static variables in question may differ between different expansions.
This is a manual merge and adaptation of pull request #P95 by @bschommer.
|
|
|
|
|
|
|
|
|
|
|
|
| |
Before, we would pass just the `ctx_vararg` component of the context
to `elab_expr` as a Boolean argument.
For future extensions, we will need to pass more of the context to
`elab_expr`, so why not pass the whole context?
This is what this commit does. The `stmt_context` type is renamed
`elab_context` and its definition is moved earlier. The `ctx`
argument is passed as is from `elab_stmt` to `elab_expr`.
|
|
|
|
|
|
|
| |
The current check for redefinition is too strict, ruling out e.g.
```
typedef int t;
void f(void) { typedef char t; }
```
|
|
|
|
|
|
| |
These are extensions w.r.t. C99, not incompatible changes.
Nothing bad can happen if those C11 features are used, except making
the code incompatible with C99.
|
|
|
|
| |
Consistently with _Noreturn, anonymous structs, etc.
|
|
|
|
|
|
| |
As the standard says (and is already implemented) an _Alignas(0)
does not change the alignment at all. The same holds for the gcc
attribute.
Bug 23387
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Allow strings literals as lvalues.
Strings and WStrings literals are lvalues, thus it is allowed to take their
addresses.
Bug 23356.
* String literals have types "array of (wide) char", not "pointer to (wide) char"
The pointer types were a leftover from the early, CIL-based C frontend.
* Remove special case for sizeof("string literal") during elaboration
No longer needed now that literals have array types.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is a follow-up to the introduction of the Storage_auto storage class
in commit 760e422.
For functions declared within a block, we used to set their storage class
to "extern":
{ {
int f(void); ---> extern int f(void);
} }
This helped enter_or_refine_ident understand that those declarations
have linkage.
Now that we have Storage_auto, this commit teaches enter_or_refine_ident
that local declarations have no linkage if auto or static, and have
linkage if extern or default.
Then, there is no need to change storage class to extern for
locally-declared functions.
In turn, this improves the "extern-after-definition" warning recently
introduced, avoiding a bizarre warning in the following case:
int foo(void){return 0;}
int main(void){
int foo(void);
return foo();
}
|
|
|
|
|
|
|
|
|
| |
Warning for change of storage class after the definition of a function
from default storage class to extern storage class. This only plays a
role if the function is also declared inline, since for inline functions
with default storage class no code is generated, but for inline functions
with extern storage class code should be generated.
Bug 23512
|
|
|
|
| |
This is what ISO C99 says, even though C++ and some C compilers accept it.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In rare cases we can end up querying the attributes of a struct, union
or enum type that is no longer in scope and therefore not bound
in the current environment.
Instead of raising an Env.Error_ exception in this case,
this commit treats such structs / unions / enums as having no
attributes attached to their definitions: only the attributes
carried by the type expression are returned.
One example where this occurs is the following, made possible by
the previous commit ("Revised elaboration of function definitions"):
int f(struct s { int a; } * p) { return p-> a; }
int g(void) { return f(NULL); }
"struct s" is scoped within the definition of f. When we get to
checking the call to f from g, checking that the NULL argument
is compatible with the "struct s *" parameters, we're outside
the scope of "struct s" and its definition is not found in the
current environment.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Change elab_type_declarator and elab_fundef_name so that the latter
returns two environments: the first one takes into account
struct/union definitions from the function return type,
while the second one also contains bindings for function
parameters and struct/union definitions from the function parameter list.
To this end the "kr_ok" bool argument of elab_type_declarator is
repurposed and renamed "fundef". It controls not just whether
K&R function declarators are supported, but also which bindings
the returned environment contains.
Change elab_fundef to adapt to the changes in elab_fundef_name
and to maintain two environments:
- the global environment, enriched with struct/union definitions from
the function return type, and with the function binding itself;
- the local environment, used for elaborating the body of the
function, which also contains bindings for function parameters
and struct/union definitions from the function parameter list.
Change elab_funbody so that it does not open a new scope for elaborating
the body, even though the body is represented as a block in the AST.
The standard says that the function body is processed in the same
scope where function parameters are declared, so that the following
is illegal:
int f(int x) { double x; ... }
Introduce a variant enter_or_refine_function of enter_or_refine_ident
where the fresh identifier to use (if no earlier declaration is found)
is created in advance in an earlier scope. This helps implement the
proper scoping of function names.
|