[LTP] [PATCH v1] Add utime07 test

[Petr Vorel]

Hi Andrea,

> This test has been extracted from symlink01 test and it verifies that
> utime() is working correctly on symlink() generated files.

According to VFS doc [1] VFS does open(2), stat(2), read(2), write(2), chmod(2).
Although the main comparison is done through stat(), it's about utime on
symlink.  struct inode_operations described in the docs contain also symlink and
update_time callbacks (filesystem specific). Wouldn't it make sense to run this
on all_filesystems?

[1] https://www.kernel.org/doc/html/latest/filesystems/vfs.html

..
> +++ b/runtest/smoketest
> @@ -9,7 +9,7 @@ wait02 wait02
>  write01 write01
>  symlink01 symlink01
>  stat04 symlink01 -T stat04
> -utime01A symlink01 -T utime01
> +utime07 utime07
>  rename01A symlink01 -T rename01
>  splice02 splice02 -s 20
>  df01_sh df01.sh
> diff --git a/runtest/syscalls b/runtest/syscalls
> index 6e2407879..5fd107e35 100644
> --- a/runtest/syscalls
> +++ b/runtest/syscalls
> @@ -1676,12 +1676,12 @@ ustat01 ustat01
>  ustat02 ustat02

>  utime01 utime01
> -utime01A symlink01 -T utime01
>  utime02 utime02
>  utime03 utime03
>  utime04 utime04
>  utime05 utime05
>  utime06 utime06
> +utime07 utime07

>  utimes01 utimes01

> diff --git a/testcases/kernel/syscalls/utime/.gitignore b/testcases/kernel/syscalls/utime/.gitignore
> index 94c0ae07c..403764521 100644
> --- a/testcases/kernel/syscalls/utime/.gitignore
> +++ b/testcases/kernel/syscalls/utime/.gitignore
> @@ -4,3 +4,4 @@
>  /utime04
>  /utime05
>  /utime06
> +/utime07
> diff --git a/testcases/kernel/syscalls/utime/utime07.c b/testcases/kernel/syscalls/utime/utime07.c
> new file mode 100644
> index 000000000..eaf832099
> --- /dev/null
> +++ b/testcases/kernel/syscalls/utime/utime07.c
> @@ -0,0 +1,78 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * Copyright (c) 2000 Silicon Graphics, Inc.  All Rights Reserved.
> + *    Author: David Fenner
> + *    Copilot: Jon Hendrickson

nit: original test is v2 only. But we are writing from scratch, using a previous
just as an inspiration. I suppose we still need to use the original copyright
and likely also the license v2 only.


> + * Copyright (C) 2024 Andrea Cervesato andrea.cervesato@suse.com
> + */
> +
> +/*\
> + * [Description]
> + *
> + * This test verifies that utime() is working correctly on symlink()
> + * generated files.
> + */
> +
> +#include <utime.h>
> +#include "tst_test.h"
> +
> +static void test_utime(void)
> +{
> +	char *symname = "my_symlink0";
> +	struct stat oldsym_stat;
> +	struct stat newsym_stat;
> +
> +	SAFE_SYMLINK(tst_get_tmpdir(), symname);
Original test actually also perform lstat() on the symlink
and checks for S_IFLNK. You consider SAFE_SYMLINK() enough?

The rest LGTM.

Reviewed-by: Petr Vorel <pvorel@suse.cz>

Kind regards,
Petr

> +	SAFE_STAT(symname, &oldsym_stat);
> +
> +	struct utimbuf utimes = {
> +		.actime = oldsym_stat.st_atime + 100,
> +		.modtime = oldsym_stat.st_mtime + 100
> +	};
> +
> +	TST_EXP_PASS(utime(symname, &utimes));
> +	SAFE_STAT(symname, &newsym_stat);
> +
> +	time_t temp, diff;
> +
> +	temp = newsym_stat.st_atime - oldsym_stat.st_atime;
> +	diff = newsym_stat.st_mtime - oldsym_stat.st_mtime - temp;
> +
> +	TST_EXP_EQ_LI(diff, 0);
> +
> +	SAFE_UNLINK(symname);
> +}
> +
> +static void test_utime_no_path(void)
> +{
> +	char *symname = "my_symlink1";
> +	struct utimbuf utimes;
> +
> +	SAFE_SYMLINK("bc+eFhi!k", symname);
> +	TST_EXP_FAIL(utime(symname, &utimes), ENOENT);
> +
> +	SAFE_UNLINK(symname);
> +}
> +
> +static void test_utime_loop(void)
> +{
> +	char *symname = "my_symlink2";
> +	struct utimbuf utimes;
> +
> +	SAFE_SYMLINK(symname, symname);
> +	TST_EXP_FAIL(utime(symname, &utimes), ELOOP);
> +
> +	SAFE_UNLINK(symname);
> +}
> +
> +static void run(void)
> +{
> +	test_utime();
> +	test_utime_no_path();
> +	test_utime_loop();
> +}
> +
> +static struct tst_test test = {
> +	.test_all = run,
> +	.needs_tmpdir = 1,
> +};