| Message ID | 20170804001639.4823-1-luc.vanoostenryck@gmail.com (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
Please ignore this patch as I must have forgot to recompile the last change. The problem is there and the solution too but bad_ctype won't solve the issue here. I'll send another version which will effectively solve the issue but I'm far from sure we'll want it. -- Luc -- To unsubscribe from this list: send the line "unsubscribe linux-sparse" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, Aug 4, 2017 at 8:31 AM, Luc Van Oostenryck <luc.vanoostenryck@gmail.com> wrote: > Please ignore this patch as I must have forgot to recompile the last change. > The problem is there and the solution too but bad_ctype won't solve > the issue here. > > I'll send another version which will effectively solve the issue but > I'm far from sure > we'll want it. > > -- Luc > -- > To unsubscribe from this list: send the line "unsubscribe linux-sparse" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-sparse" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, Aug 4, 2017 at 8:31 AM, Luc Van Oostenryck <luc.vanoostenryck@gmail.com> wrote: > Please ignore this patch as I must have forgot to recompile the last change. > The problem is there and the solution too but bad_ctype won't solve > the issue here. > > I'll send another version which will effectively solve the issue but > I'm far from sure > we'll want it. Sorry I hit the send before I type up some reply in my last email. I will wait for your new patch then. Chris -- To unsubscribe from this list: send the line "unsubscribe linux-sparse" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/evaluate.c b/evaluate.c index cf3cf244d..5b4abdb6a 100644 --- a/evaluate.c +++ b/evaluate.c @@ -1220,6 +1220,7 @@ static struct symbol *evaluate_conditional_expression(struct expression *expr) Err: expression_error(expr, "incompatible types in conditional expression (%s)", typediff); + expr->ctype = &bad_ctype; return NULL; out: diff --git a/validation/cond-err-expand.c b/validation/cond-err-expand.c new file mode 100644 index 000000000..72af8d4b1 --- /dev/null +++ b/validation/cond-err-expand.c @@ -0,0 +1,27 @@ +static inline void f(void) +{ + __builtin_constant_p(0); +} + +int foo(int a) +{ + return 0 ? 0 : f(); +} + +int bar(int a) +{ + return 1 ? f() : 0; +} + +/* + * check-name: cond-err-expand.c + * check-command: test-linearize -Wno-decl $file + * + * check-error-start +cond-err-expand.c:8:18: error: incompatible types in conditional expression (different base types) +cond-err-expand.c:13:18: error: incompatible types in conditional expression (different base types) + * check-error-end + * + * check-output-ignore + * check-excludes: call.* __builtin_constant_p + */
Bad conditional expressions used to have no type (NULL) this in turn makes that some further processing are not done. In particular, here, the expansion of the operands are not done. Fix this by giving to such expression a type 'bad_type'. Note: nor gcc, not clang seems to emit a warning for the the testcase here which is not conform to the standard. OTOH, sparse complains and this was the cause of the non-expansion of the builtin. Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com> --- evaluate.c | 1 + validation/cond-err-expand.c | 27 +++++++++++++++++++++++++++ 2 files changed, 28 insertions(+) create mode 100644 validation/cond-err-expand.c