From: Petr Vorel <pvorel@suse.cz>
To: Wei Gao <wegao@suse.com>
Cc: ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v3 2/2] getcwd01: Implement .test_variants
Date: Tue, 23 Jan 2024 10:45:54 +0100 [thread overview]
Message-ID: <20240123094554.GC149835@pevik> (raw)
In-Reply-To: <20240117125227.24700-3-wegao@suse.com>
Hi Wei,
> Signed-off-by: Wei Gao <wegao@suse.com>
> ---
> testcases/kernel/syscalls/getcwd/getcwd01.c | 80 +++++++++++++++------
> 1 file changed, 60 insertions(+), 20 deletions(-)
> NOTE: Cyril give solution for run the test in a child and pass
> the test both on EFAULT and child being killed by SIGSEGV. But
> currently i have no idea how to do it since no SIGSEGV will hapeen
> if NULL buffer give to getcwd. This file just used for give a real
> user case for TST_EXP_FAIL_PTR.
@Cyril, could you please point out test which uses similar approach?
Maybe these?
testcases/kernel/syscalls/fstat/fstat03.c
testcases/kernel/syscalls/ipc/shmat/shmat01.c
testcases/kernel/syscalls/kill/kill11.c
testcases/kernel/syscalls/setrlimit/setrlimit05.c
Or can the current approach be used?
Below only nits (formatting).
> diff --git a/testcases/kernel/syscalls/getcwd/getcwd01.c b/testcases/kernel/syscalls/getcwd/getcwd01.c
> index 218bf4ef2..879c36206 100644
> --- a/testcases/kernel/syscalls/getcwd/getcwd01.c
> +++ b/testcases/kernel/syscalls/getcwd/getcwd01.c
> @@ -3,21 +3,34 @@
> * Copyright (c) International Business Machines Corp., 2001
IBM in 2001 could be proud on this code :). There should have been some
copyright, I would add:
* Copyright (c) Linux Test Project, 2002-2024
> */
> -/*
> - * DESCRIPTION
> +/*\
> + * [Description]
> + *
> * Testcase to test that getcwd(2) sets errno correctly.
> + *
> + * 1. getcwd(2) fails if buf points to a bad address.
> + * 2. getcwd(2) fails if the size is invalid.
> + * 3. getcwd(2) fails if the size is set to 0.
> + * 4. getcwd(2) fails if the size is set to 1.
> + * 5. getcwd(2) fails if buf points to NULL and the size is set to 1.
> *
> * Expected Result:
> + *
> + * linux syscall
> + *
> + * 1. getcwd(2) should return NULL and set errno to EFAULT.
I don't like repeating "getcwd(2) should return NULL and set errno to"
> + * 2. getcwd(2) should return NULL and set errno to EFAULT.
> + * 3. getcwd(2) should return NULL and set errno to ERANGE.
> + * 4. getcwd(2) should return NULL and set errno to ERANGE.
> + * 5. getcwd(2) should return NULL and set errno to ERANGE.
> + *
> + * glibc and uclibc{,-ng}.
Although in the past LTP developers cared only about glibc, now we generally
don't stick to any libc implementation. Thus it should be "libc syscall wrapper"
> + *
> + * 1. getcwd(2) should return NULL and set errno to EFAULT.
> + * 2. getcwd(2) should return NULL and set errno to ENOMEM.
> + * 3. getcwd(2) should return NULL and set errno to EINVAL.
> + * 4. getcwd(2) should return NULL and set errno to ERANGE.
> + * 5. getcwd(2) should return NULL and set errno to ERANGE.
> */
How about to make it simple like this:
/*\
* [Description]
*
* Testcase to test that getcwd(2) returns NULL and sets errno correctly.
*
* 1. getcwd(2) fails if buf points to a bad address (EFAULT).
* 2. getcwd(2) fails if the size is invalid (syscall: EFAULT, libc wrapper: ENOMEM).
* 3. getcwd(2) fails if the size is set to 0 (syscall: ERANGE, libc wrapper: EINVAL).
* 4. getcwd(2) fails if the size is set to 1 (ERANGE).
* 5. getcwd(2) fails if buf points to NULL and the size is set to 1 (ERANGE).
*/
> #include <errno.h>
> @@ -32,23 +45,50 @@ static struct t_case {
> char *buf;
> size_t size;
> int exp_err;
> + int exp_err_libc;
> } tcases[] = {
> - {(void *)-1, PATH_MAX, EFAULT},
> - {NULL, (size_t)-1, EFAULT},
> - {buffer, 0, ERANGE},
> - {buffer, 1, ERANGE},
> - {NULL, 1, ERANGE}
> + {(void *)-1, PATH_MAX, EFAULT, EFAULT},
> + {NULL, (size_t)-1, EFAULT, ENOMEM},
> + {buffer, 0, ERANGE, EINVAL},
> + {buffer, 1, ERANGE, ERANGE},
> + {NULL, 1, ERANGE, ERANGE},
> };
> +static inline void tst_getcwd(char *buf, size_t size, int exp_err, int exp_err_libc)
> +{
> +
> + if (tst_variant == 0)
> + TST_EXP_FAIL2(tst_syscall(__NR_getcwd, buf, size), exp_err);
> + else
> + TST_EXP_FAIL_PTR(getcwd(buf, size), exp_err_libc);
> +}
+1
> -static void verify_getcwd(unsigned int n)
> +static void run(unsigned int n)
> {
> struct t_case *tc = &tcases[n];
> - TST_EXP_FAIL2(tst_syscall(__NR_getcwd, tc->buf, tc->size), tc->exp_err);
> + /* https://github.com/linux-test-project/ltp/issues/1084 */
IMHO this comment should go to the commit message.
> +#if !defined(__GLIBC__) && !defined(__ANDROID__)
I'll ask AOSP developers to check this.
> + if (tst_variant && !tc->buf) {
> + tst_res(TCONF, "NULL buffer test skipped on MUSL due different implementation");
> + return;
> + }
> +#endif
> +
> + tst_getcwd(tc->buf, tc->size, tc->exp_err, tc->exp_err_libc);
> +}
> +
> +static void setup(void)
> +{
> + if (tst_variant == 0)
> + tst_res(TINFO, "Testing getcwd with raw syscall");
> + else
> + tst_res(TINFO, "Testing getcwd with wrap syscall");
"wrap syscall" => "libc syscall wrapper"
Kind regards,
Petr
--
Mailing list info: https://lists.linux.it/listinfo/ltp
next prev parent reply other threads:[~2024-01-23 9:46 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-11 1:26 [LTP] [PATCH v1] Add TST_EXP_FAIL_PTR Wei Gao via ltp
2024-01-16 17:49 ` Petr Vorel
2024-01-17 8:07 ` Wei Gao via ltp
2024-01-17 8:04 ` [LTP] [PATCH v2] " Wei Gao via ltp
2024-01-17 9:49 ` Petr Vorel
2024-01-30 12:21 ` Petr Vorel
2024-01-17 12:52 ` [LTP] [PATCH v3 0/2] lib: TST_EXP_FAIL_PTR Wei Gao via ltp
2024-01-17 12:52 ` [LTP] [PATCH v3 1/2] " Wei Gao via ltp
2024-01-23 8:45 ` Petr Vorel
2024-01-23 10:41 ` Cyril Hrubis
2024-01-30 12:20 ` Petr Vorel
2024-01-17 12:52 ` [LTP] [PATCH v3 2/2] getcwd01: Implement .test_variants Wei Gao via ltp
2024-01-23 9:45 ` Petr Vorel [this message]
2024-02-08 1:32 ` [LTP] [PATCH v4 0/3] lib: TST_EXP_{FAIL,PASS}_PTR_{NULL,VOID} Wei Gao via ltp
2024-02-08 1:32 ` [LTP] [PATCH v4 1/3] " Wei Gao via ltp
2024-03-20 10:47 ` Petr Vorel
2024-03-26 10:54 ` Cyril Hrubis
2024-02-08 1:32 ` [LTP] [PATCH v4 2/3] shmat02.c: Use TST_EXP_FAIL_PTR_VOID Wei Gao via ltp
2024-03-20 10:47 ` Petr Vorel
2024-02-08 1:32 ` [LTP] [PATCH v4 3/3] realpath01.c: use TST_EXP_FAIL_PTR_NULL Wei Gao via ltp
2024-03-20 10:47 ` Petr Vorel
2024-03-27 3:49 ` [LTP] [PATCH v5 0/3] lib: TST_EXP_{FAIL,PASS}_PTR_{NULL,VOID} Wei Gao via ltp
2024-03-27 3:49 ` [LTP] [PATCH v5 1/3] " Wei Gao via ltp
2024-03-28 11:29 ` Petr Vorel
2024-03-28 11:49 ` Petr Vorel
2024-03-28 11:57 ` Petr Vorel
2024-03-28 12:11 ` Petr Vorel
2024-03-27 3:49 ` [LTP] [PATCH v5 2/3] shmat02.c: Use TST_EXP_FAIL_PTR_VOID Wei Gao via ltp
2024-03-27 3:49 ` [LTP] [PATCH v5 3/3] realpath01.c: use TST_EXP_FAIL_PTR_NULL Wei Gao via ltp
2024-04-03 3:28 ` [LTP] [PATCH v6 0/3] lib: TST_EXP_{FAIL,PASS}_PTR_{NULL,VOID} Wei Gao via ltp
2024-04-03 3:28 ` [LTP] [PATCH v6 1/3] " Wei Gao via ltp
2024-04-04 13:51 ` Cyril Hrubis
2024-04-04 16:01 ` Petr Vorel
2024-04-05 8:53 ` Cyril Hrubis
2024-04-05 10:28 ` Petr Vorel
2024-04-05 11:23 ` Cyril Hrubis
2024-04-05 14:03 ` Petr Vorel
2024-04-07 23:31 ` Wei Gao via ltp
2024-04-03 3:28 ` [LTP] [PATCH v6 2/3] shmat02.c: Use TST_EXP_FAIL_PTR_VOID Wei Gao via ltp
2024-04-03 3:28 ` [LTP] [PATCH v6 3/3] realpath01.c: use TST_EXP_FAIL_PTR_NULL 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=20240123094554.GC149835@pevik \
--to=pvorel@suse.cz \
--cc=ltp@lists.linux.it \
--cc=wegao@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