Linux Test Project
 help / color / mirror / Atom feed
From: Andrea Cervesato via ltp <ltp@lists.linux.it>
To: "Wei Gao via ltp" <ltp@lists.linux.it>
Cc: ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v9] connect03: New test case for EPROTOTYPE and EACCES errors
Date: Thu, 09 Jul 2026 16:43:06 +0000	[thread overview]
Message-ID: <6a4fcf9c.795e137c.249a98.9ebd@mx.google.com> (raw)
In-Reply-To: <20260706234000.23719-1-wegao@suse.com>

Hi Wei,

we can remove privileged user needs for this test.

> +static int sys_connect(int sockfd, const struct sockaddr *addr,
> +		       socklen_t addrlen)
> +{
> +	return tst_syscall(__NR_connect, sockfd, addr, addrlen);
> +}

we should probably use connect() here instead of raw syscall.

> +
> +static void setup(void)
> +{
> +	sock_un.sun_family = AF_UNIX;
> +	strncpy(sock_un.sun_path, SOCK_FILE, sizeof(sock_un.sun_path));
> +
> +	fd_unix_server = SAFE_SOCKET(AF_UNIX, SOCK_STREAM, 0);
> +	SAFE_BIND(fd_unix_server, (struct sockaddr *)&sock_un, sizeof(sock_un));
> +	SAFE_CHMOD(SOCK_FILE, 0700);
> +	SAFE_LISTEN(fd_unix_server, 5);
> +
> +	fd_unix_dgram = SAFE_SOCKET(AF_UNIX, SOCK_DGRAM, 0);
> +	fd_unix_stream = SAFE_SOCKET(AF_UNIX, SOCK_STREAM, 0);
> +
> +	pw = SAFE_GETPWNAM("nobody");

we only need this to trigger EACCES, so probably chmod(0500) is enough.
Read below.

> +}
> +
> +static void cleanup(void)
> +{
> +	if (fd_unix_dgram != -1)
> +		SAFE_CLOSE(fd_unix_dgram);
> +	if (fd_unix_stream != -1)
> +		SAFE_CLOSE(fd_unix_stream);
> +	if (fd_unix_server != -1)
> +		SAFE_CLOSE(fd_unix_server);
> +}
> +
> +static void verify_connect(unsigned int i)
> +{
> +	struct test_case_t *tc = &tcases[i];
> +
> +	if (tc->exp_errno == EACCES) {
> +		if (!SAFE_FORK()) {
> +			SAFE_SETUID(pw->pw_uid);
> +			TST_EXP_FAIL(sys_connect(*tc->fd, (const struct sockaddr *)tc->addr, tc->salen),
> +				     tc->exp_errno, "%s", tc->desc);
> +			exit(0);
> +		}
> +		tst_reap_children();
> +	} else {
> +		TST_EXP_FAIL(sys_connect(*tc->fd, (const struct sockaddr *)tc->addr, tc->salen),
> +			     tc->exp_errno, "%s", tc->desc);
> +	}

This can easily become:

	SAFE_CHMOD(SOCK_FILE, tc->mode);

	TST_EXP_FAIL(connect(*tc->fd, (struct sockaddr *)&sock_un, sizeof(sock_un)),
		     tc->exp_errno, "%s", tc->desc);

To make it works, we need a mode attribute in the tcase struct:

- 0700 for EPROTOTYPE
- 0500 for EACCES

> +}
> +
> +static struct tst_test test = {
> +	.setup = setup,
> +	.cleanup = cleanup,
> +	.tcnt = ARRAY_SIZE(tcases),
> +	.test = verify_connect,
> +	.forks_child = 1,

this is not needed anymore.

> +	.needs_root = 1,

And root can go away.

> +	.needs_tmpdir = 1,
> +};
> -- 
> 2.54.0
> 
> 
> -- 
> Mailing list info: https://lists.linux.it/listinfo/ltp

Regards,
--
Andrea Cervesato
SUSE QE Automation Engineer Linux
andrea.cervesato@suse.com

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

  parent reply	other threads:[~2026-07-09 16:43 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-14  4:51 [LTP] [PATCH v2 1/2] connect01: Convert to new API Yang Xu via ltp
2024-05-14  4:51 ` [LTP] [PATCH v2 2/2] connect01: Add negative tests Yang Xu via ltp
2024-07-17 12:30 ` [LTP] [PATCH v2 1/2] connect01: Convert to new API Cyril Hrubis
2026-05-25  9:02 ` [LTP] [PATCH v3 0/2] " Wei Gao via ltp
2026-05-25  9:02   ` [LTP] [PATCH v3 1/2] " Wei Gao via ltp
2026-05-25 10:12     ` [LTP] " linuxtestproject.agent
2026-05-28  4:43       ` Wei Gao via ltp
2026-06-04  4:49     ` [LTP] [PATCH v4 0/2] " Wei Gao via ltp
2026-06-04  4:49       ` [LTP] [PATCH v4 1/2] " Wei Gao via ltp
2026-06-04  7:15         ` [LTP] " linuxtestproject.agent
2026-06-16  1:47         ` [LTP] [PATCH v5 0/2] " Wei Gao via ltp
2026-06-16  1:47           ` [LTP] [PATCH v5 1/2] " Wei Gao via ltp
2026-06-16  4:07             ` [LTP] " linuxtestproject.agent
2026-06-16  5:24             ` [LTP] [PATCH v6 0/2] " Wei Gao via ltp
2026-06-16  5:24               ` [LTP] [PATCH v6 1/2] " Wei Gao via ltp
2026-06-16  8:35                 ` [LTP] " linuxtestproject.agent
2026-06-16  9:31                 ` [LTP] [PATCH v7 0/2] " Wei Gao via ltp
2026-06-16  9:31                   ` [LTP] [PATCH v7 1/2] " Wei Gao via ltp
2026-06-16 10:12                     ` [LTP] " linuxtestproject.agent
2026-07-01  8:34                     ` [LTP] [PATCH v7 1/2] " Petr Vorel
2026-06-16  9:31                   ` [LTP] [PATCH v7 2/2] connect01: Add negative tests Wei Gao via ltp
2026-07-01  8:38                     ` Petr Vorel
2026-07-06  6:02                     ` [LTP] [PATCH v8] connect03: New test case for EPROTOTYPE and EACCES errors Wei Gao via ltp
2026-07-06  8:22                       ` [LTP] " linuxtestproject.agent
2026-07-06 23:39                       ` [LTP] [PATCH v9] " Wei Gao via ltp
2026-07-07  3:25                         ` [LTP] " linuxtestproject.agent
2026-07-09 16:43                         ` Andrea Cervesato via ltp [this message]
2026-07-10  5:13                         ` [LTP] [PATCH v10] " Wei Gao via ltp
2026-07-10  8:31                           ` [LTP] " linuxtestproject.agent
2026-07-13  7:52                           ` [LTP] [PATCH v10] " Andrea Cervesato via ltp
2026-06-16  5:24               ` [LTP] [PATCH v6 2/2] connect01: Add negative tests Wei Gao via ltp
2026-06-16  1:47           ` [LTP] [PATCH v5 " Wei Gao via ltp
2026-06-04  4:49       ` [LTP] [PATCH v4 " Wei Gao via ltp
2026-05-25  9:02   ` [LTP] [PATCH v3 " Wei Gao via ltp

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=6a4fcf9c.795e137c.249a98.9ebd@mx.google.com \
    --to=ltp@lists.linux.it \
    --cc=andrea.cervesato@suse.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox