From mboxrd@z Thu Jan  1 00:00:00 1970
From: Cyril Hrubis <chrubis@suse.cz>
Date: Thu, 22 Jul 2021 14:08:20 +0200
Subject: [LTP] [PATCH v1 4/4] syscalls/shmget06: Add test when the id of
 shm_next_id has existed
In-Reply-To: <1620809541-6891-4-git-send-email-xuyang2018.jy@fujitsu.com>
References: <1620809541-6891-1-git-send-email-xuyang2018.jy@fujitsu.com>
 <1620809541-6891-4-git-send-email-xuyang2018.jy@fujitsu.com>
Message-ID: <YPlftOtF7QxjjDnX@yuki>
List-Id: <ltp.lists.linux.it>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: ltp@lists.linux.it

Hi!
> diff --git a/testcases/kernel/syscalls/ipc/shmget/shmget06.c b/testcases/kernel/syscalls/ipc/shmget/shmget06.c
> new file mode 100644
> index 000000000..3138d4482
> --- /dev/null
> +++ b/testcases/kernel/syscalls/ipc/shmget/shmget06.c
> @@ -0,0 +1,77 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * Copyright (c) 2021 FUJITSU LIMITED. All rights reserved.
> + * Author: Yang Xu <xuyang2018.jy@fujitsu.com>
> + */
> +
> +/*\
> + * [Description]
> + *
> + * It is a basic test about shm_next_id.
                          ^
			  for
> + *
> + * When the shared memory segment identifier that shm_next_id stored has
> + * existed, call shmget with different key just use another unused value in range
      ^                                      ^
      'does exist' instead of 'has existed'  |
      or even better 'is allready in use'    |
                                            will
> + * [0,INT_MAX]. kernel doesn't guarantee the desired id.
                   ^
		   Capital letter at the start of the sentence.
> + */
> +
> +#include <errno.h>
> +#include <string.h>
> +#include <sys/types.h>
> +#include <sys/ipc.h>
> +#include <sys/shm.h>
> +#include "tst_test.h"
> +#include "tst_safe_sysv_ipc.h"
> +#include "libnewipc.h"
> +
> +#define NEXT_ID_PATH "/proc/sys/kernel/shm_next_id"
> +
> +static int shm_id[2], pid;
> +static key_t shmkey[2];
> +
> +static void verify_shmget(void)
> +{
> +	SAFE_FILE_PRINTF(NEXT_ID_PATH, "%d", shm_id[0]);
> +
> +	shm_id[1] = SAFE_SHMGET(shmkey[1], SHM_SIZE, IPC_CREAT | SHM_RW);
> +	if (shm_id[1] == shm_id[0])
> +		tst_res(TFAIL, "shm id %d has existed, shmget() returns the"
> +			" same shm id unexpectedly", shm_id[0]);
> +	else
> +		tst_res(TPASS, "shm id %d has existed, shmget() returns the"
> +			" new shm id %d", shm_id[0], shm_id[1]);
> +
> +	SAFE_SHMCTL(shm_id[1], IPC_RMID, NULL);
> +}
> +
> +static void setup(void)
> +{
> +	shmkey[0] = GETIPCKEY();
> +	shmkey[1] = GETIPCKEY();
> +	pid = getpid();
> +	SAFE_FILE_PRINTF(NEXT_ID_PATH, "%d", pid);

So we are using this to "randomize" the id here right?

> +	shm_id[0] = SAFE_SHMGET(shmkey[0], SHM_SIZE, IPC_CREAT | SHM_RW);
> +	tst_res(TINFO, "Test shm_next_id effects on shmget(different key) "
> +		"when this identifier of shared memory segment has existed");

I do not think that printing test description is useful, especially
since we have nicely formatted test description in the test source which
gets exported into the metadata.

> +}
> +
> +static void cleanup(void)
> +{
> +	int i;
> +
> +	for (i = 0; i < 2; i++) {
> +		if (shm_id[i] != -1)
> +			SAFE_SHMCTL(shm_id[i], IPC_RMID, NULL);
> +	}
> +}
> +
> +static struct tst_test test = {
> +	.needs_tmpdir = 1,
> +	.setup = setup,
> +	.cleanup = cleanup,
> +	.test_all = verify_shmget,
> +	.needs_kconfigs = (const char *[]) {
> +		"CONFIG_CHECKPOINT_RESTORE=y",
> +		NULL
> +	},
> +	.needs_root = 1,
> +};

With the minor adjustements in description:

Reivewed-by: Cyril Hrubis <chrubis@suse.cz>

-- 
Cyril Hrubis
chrubis@suse.cz