The branch main has been updated by glebius:
URL:
https://cgit.FreeBSD.org/src/commit/?id=3ad0f9a584106b85569c6366ef8117a664fd55f8
commit 3ad0f9a584106b85569c6366ef8117a664fd55f8
Author: Gleb Smirnoff <gleb...@freebsd.org>
AuthorDate: 2024-02-08 17:00:23 +0000
Commit: Gleb Smirnoff <gleb...@freebsd.org>
CommitDate: 2024-02-08 17:00:41 +0000
tests/unix_passfd: add a comment for send_a_lot
Explain why SOCK_DGRAM and SOCK_STREAM differ in this edge case. While
here improve output in case of a failure.
---
tests/sys/kern/unix_passfd_test.c | 13 +++++++++++--
1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/tests/sys/kern/unix_passfd_test.c
b/tests/sys/kern/unix_passfd_test.c
index 143ccd098583..74095859d899 100644
--- a/tests/sys/kern/unix_passfd_test.c
+++ b/tests/sys/kern/unix_passfd_test.c
@@ -657,11 +657,20 @@ ATF_TC_BODY(rights_creds_payload, tc)
ATF_REQUIRE_MSG(len != -1 , "sendmsg failed: %s", strerror(errno));
#if TEST_PROTO == SOCK_STREAM
ATF_REQUIRE_MSG((size_t)len < sendspace,
- "sendmsg: %zd bytes sent", len);
+ "sendmsg: %zd bytes sent, expected < %lu", len, sendspace);
#endif
#if TEST_PROTO == SOCK_DGRAM
+ /*
+ * sendmsg(2) can't truncate datagrams, only recvmsg(2) can. There are
+ * two options for the kernel here: either accept the datagram with
+ * slight overcommit of the socket buffer space or return ENOBUFS for a
+ * datagram that is smaller or equal to the socket buffer space. Our
+ * implementation does overcommit. Explanation is simple: from our
+ * side we see space available, we have no idea that remote side has
+ * LOCAL_CREDS set. From our side we expect sendmsg(2) to succeed.
+ */
ATF_REQUIRE_MSG((size_t)len == sendspace,
- "sendmsg: %zd bytes sent", len);
+ "sendmsg: %zd bytes sent, expected %lu", len, sendspace);
#endif
rlen = recvfd_payload(fd[1], &getfd, buf, len,
CMSG_SPACE(SOCKCREDSIZE(CMGROUP_MAX)) + CMSG_SPACE(sizeof(int)), 0);
