Good evening,
This is regarding a regression detected on Linaro here:
https://gcc.gnu.org/pipermail/gcc-regression/2026-September/084452.html
It now occurred to me that 'test' in "test for excess errors"
might be a imperative verb, and not a noun. It would make sense,
since now this test is emitting two errors instead of one (because
we now bail early).
All the detected regressions are 'test for excess errors' on this
line. One is C, other three are different standards of C++.
Let me know,
Léo
--------------
Fixes a regression in r17-4537-g059231fb8a42ea.
A test that only produced one error now outputs two.
gcc/testsuite/ChangeLog:
* c-c++-common/gomp/has-include-1.c: Change from dg-error to
dg-excess-errors.
---
gcc/testsuite/c-c++-common/gomp/has-include-1.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/gcc/testsuite/c-c++-common/gomp/has-include-1.c
b/gcc/testsuite/c-c++-common/gomp/has-include-1.c
index bb1b9052685..00fcfa26f73 100644
--- a/gcc/testsuite/c-c++-common/gomp/has-include-1.c
+++ b/gcc/testsuite/c-c++-common/gomp/has-include-1.c
@@ -3,6 +3,6 @@
void
foo (void)
{
-#pragma omp parallel if (__has_include ("<stdlib.h>")) /* { dg-error "used
outside of preprocessing directive" } */
+#pragma omp parallel if (__has_include ("<stdlib.h>")) /* { dg-excess-errors
"used outside of preprocessing directive" } */
;
}
--
2.47.3
