| Message ID | 20160301141052.GZ14947@redhat.com |
|---|---|
| State | New |
| Headers | show |
On 03/01/2016 07:10 AM, Marek Polacek wrote: > This PR from 2011 reports that -Waddress prints unhelpful warning when the > comparison comes from a macro. Since I've added from_macro_expansion_at, > this is easy to circumvent. I'm not so sure we actually want to disable > the warning in the case of a macro, but probably yes. > > Bootstrapped/regtested on x86_64-linux, ok for trunk or should I defer to > GCC 7? > > 2016-03-01 Marek Polacek <polacek@redhat.com> > > PR c/48778 > * c-typeck.c (build_binary_op): Don't issue -Waddress warnings > for macro expansions. > > * gcc.dg/Waddress-2.c: New test. I think deferral is the best option here. jeff
On Tue, Mar 01, 2016 at 04:38:13PM -0700, Jeff Law wrote: > On 03/01/2016 07:10 AM, Marek Polacek wrote: > >This PR from 2011 reports that -Waddress prints unhelpful warning when the > >comparison comes from a macro. Since I've added from_macro_expansion_at, > >this is easy to circumvent. I'm not so sure we actually want to disable > >the warning in the case of a macro, but probably yes. > > > >Bootstrapped/regtested on x86_64-linux, ok for trunk or should I defer to > >GCC 7? > > > >2016-03-01 Marek Polacek <polacek@redhat.com> > > > > PR c/48778 > > * c-typeck.c (build_binary_op): Don't issue -Waddress warnings > > for macro expansions. > > > > * gcc.dg/Waddress-2.c: New test. > I think deferral is the best option here. I agree. I will hold off, thanks. Marek
On 03/01/2016 07:10 AM, Marek Polacek wrote: > This PR from 2011 reports that -Waddress prints unhelpful warning when the > comparison comes from a macro. Since I've added from_macro_expansion_at, > this is easy to circumvent. I'm not so sure we actually want to disable > the warning in the case of a macro, but probably yes. > > Bootstrapped/regtested on x86_64-linux, ok for trunk or should I defer to > GCC 7? > > 2016-03-01 Marek Polacek <polacek@redhat.com> > > PR c/48778 > * c-typeck.c (build_binary_op): Don't issue -Waddress warnings > for macro expansions. > > * gcc.dg/Waddress-2.c: New test. I've got mixed feelings about this patch. Though we have traditionally desired to suppress some warnings that occur due to macro expansions, based on that, I'll ack for the trunk. Thanks, jeff
diff --git gcc/c/c-typeck.c gcc/c/c-typeck.c index 6aa0f03..0f415e0 100644 --- gcc/c/c-typeck.c +++ gcc/c/c-typeck.c @@ -11087,7 +11087,8 @@ build_binary_op (location_t location, enum tree_code code, else if (code0 == POINTER_TYPE && null_pointer_constant_p (orig_op1)) { if (TREE_CODE (op0) == ADDR_EXPR - && decl_with_nonnull_addr_p (TREE_OPERAND (op0, 0))) + && decl_with_nonnull_addr_p (TREE_OPERAND (op0, 0)) + && !from_macro_expansion_at (location)) { if (code == EQ_EXPR) warning_at (location, @@ -11107,7 +11108,8 @@ build_binary_op (location_t location, enum tree_code code, else if (code1 == POINTER_TYPE && null_pointer_constant_p (orig_op0)) { if (TREE_CODE (op1) == ADDR_EXPR - && decl_with_nonnull_addr_p (TREE_OPERAND (op1, 0))) + && decl_with_nonnull_addr_p (TREE_OPERAND (op1, 0)) + && !from_macro_expansion_at (location)) { if (code == EQ_EXPR) warning_at (location, diff --git gcc/testsuite/gcc.dg/Waddress-2.c gcc/testsuite/gcc.dg/Waddress-2.c index e69de29..4d927f6 100644 --- gcc/testsuite/gcc.dg/Waddress-2.c +++ gcc/testsuite/gcc.dg/Waddress-2.c @@ -0,0 +1,24 @@ +/* PR c/48778 */ +/* { dg-do compile } */ +/* { dg-options "-Waddress" } */ + +#define NULL ((void *) 0) + +#define M1(b) ((b) != NULL ? 0 : (b)) +#define M2(b) ((b) == NULL ? 0 : (b)) +#define M3(b) (NULL != (b) ? 0 : (b)) +#define M4(b) (NULL == (b) ? 0 : (b)) + +int +func (int b) +{ + if (M1 (&b) > 0) + return 1; + if (M2 (&b) > 0) + return 2; + if (M3 (&b) > 0) + return 3; + if (M4 (&b) > 0) + return 4; + return 0; +}