https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79074
Bug ID: 79074
Summary: -Waddress difference between C and C++ with (T*)0
Product: gcc
Version: 7.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: msebor at gcc dot gnu.org
Target Milestone: ---
The regression test added for bug 47931 fails in C++ mode because C++ warns on
pointless comparisons with a null pointer even if it's not a null pointer
constant (e.g., (int*)0). Fixing the failure by adding the appropriate
dg-warning directive makes the test fail in C (where it runs with -Wc++-compat)
because C only warns only for comparisons with a null pointer constant. It
seems that the two front ends should warn consistently for same expressions.
If there's some reason not to warn in C when the null pointer doesn't strictly
meet the requirements of a null pointer constant then it should warn with
-Wc++-compat for compatibility.
$ (set -x; cat t.c && for l in c c++; do /build/gcc-svn/gcc/xgcc -B
/build/gcc-svn/gcc -S -Waddress -x$l t.c; done)
+ cat t.c
int f (int i) { return &i != (void *) 0; }
int g (int i) { return &i != (int *) 0; }
+ for l in c c++
+ /build/gcc-svn/gcc/xgcc -B /build/gcc-svn/gcc -S -Waddress -xc t.c
t.c: In function ‘f’:
t.c:1:27: warning: the comparison will always evaluate as ‘true’ for the
address of ‘i’ will never be NULL [-Waddress]
int f (int i) { return &i != (void *) 0; }
^~
+ for l in c c++
+ /build/gcc-svn/gcc/xgcc -B /build/gcc-svn/gcc -S -Waddress -xc++ t.c
t.c: In function ‘int f(int)’:
t.c:1:27: warning: the address of ‘i’ will never be NULL [-Waddress]
int f (int i) { return &i != (void *) 0; }
~~~^~~~~~~~~~~~~
t.c: In function ‘int g(int)’:
t.c:3:27: warning: the address of ‘i’ will never be NULL [-Waddress]
int g (int i) { return &i != (int *) 0; }
~~~^~~~~~~~~~~~
