It's possible for resolve_ref_unsafe() to return NULL (e.g.,
if we are reading and the ref does not exist), in which case
we'll pass NULL to printf. On glibc systems this produces
"(null)", but on others it may segfault.
The tests don't expect any such case, but if we ever did
trigger this, we would prefer to cleanly fail the test with
unexpected input rather than segfault. Let's manually
replace NULL with "(null)". The exact value doesn't matter,
as it won't match any possible ref the caller could expect
(and anyway, the exit code of the program will tell whether
"ref" is valid or not).
Signed-off-by: Jeff King <peff@xxxxxxxx>
---
t/helper/test-ref-store.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/t/helper/test-ref-store.c b/t/helper/test-ref-store.c
index 05d8c4d8af..6ec2670044 100644
--- a/t/helper/test-ref-store.c
+++ b/t/helper/test-ref-store.c
@@ -135,7 +135,7 @@ static int cmd_resolve_ref(struct ref_store *refs, const char **argv)
ref = refs_resolve_ref_unsafe(refs, refname, resolve_flags,
sha1, &flags);
- printf("%s %s 0x%x\n", sha1_to_hex(sha1), ref, flags);
+ printf("%s %s 0x%x\n", sha1_to_hex(sha1), ref ? ref : "(null)", flags);
return ref ? 0 : 1;
}
--
2.15.0.rc1.560.g5f0609e481
