Files
llvm/clang/test/Analysis/fields.c
Jordan Rose 12024f8776 Revert "[analyzer] Check that a member expr is valid even when the result is an lvalue."
The original intent of this commit was to catch potential null dereferences
early, but it breaks the common "home-grown offsetof" idiom (PR13927):

 (((struct Foo *)0)->member - ((struct foo *)0))

As it turns out, this appears to be legal in C, per a footnote in
C11 6.5.3.2: "Thus, &*E is equivalent to E (even if E is a null pointer)".
In C++ this issue is still open:
  http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_active.html#232

We'll just have to make sure we have good path notes in the future.

This reverts r164441 / 9be016dcd1ca3986873a7b66bd4bc027309ceb59.

llvm-svn: 164958
2012-10-01 19:07:15 +00:00

29 lines
398 B
C

// RUN: %clang_cc1 -analyze -analyzer-checker=core,alpha.core %s -analyzer-store=region -verify
unsigned foo();
typedef struct bf { unsigned x:2; } bf;
void bar() {
bf y;
*(unsigned*)&y = foo();
y.x = 1;
}
struct s {
int n;
};
void f() {
struct s a;
int *p = &(a.n) + 1;
}
typedef struct {
int x,y;
} Point;
Point getit(void);
void test() {
Point p;
(void)(p = getit()).x;
}