Re: [LTP] [PATCH] setitimer01: rewrite using new API

[Richard Palethorpe <rpalethorpe@suse.de>]

Hello,

Li Wang <liwang@redhat.com> writes:

> Also add signal checking when the timer take effection.
>
> Signed-off-by: Li Wang <liwang@redhat.com>
> ---
>  .../kernel/syscalls/setitimer/setitimer01.c   | 230 ++++++++----------
>  1 file changed, 99 insertions(+), 131 deletions(-)
>
> diff --git a/testcases/kernel/syscalls/setitimer/setitimer01.c b/testcases/kernel/syscalls/setitimer/setitimer01.c
> index 6874b94ad..def559db8 100644
> --- a/testcases/kernel/syscalls/setitimer/setitimer01.c
> +++ b/testcases/kernel/syscalls/setitimer/setitimer01.c
> @@ -1,157 +1,125 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
>  /*
> + * Copyright (c) International Business Machines  Corp., 2001
> + * 03/2001 - Written by Wayne Boyer
>   *
> - *   Copyright (c) International Business Machines  Corp., 2001
> - *
> - *   This program is free software;  you can redistribute it and/or modify
> - *   it under the terms of the GNU General Public License as published by
> - *   the Free Software Foundation; either version 2 of the License, or
> - *   (at your option) any later version.
> - *
> - *   This program is distributed in the hope that it will be useful,
> - *   but WITHOUT ANY WARRANTY;  without even the implied warranty of
> - *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See
> - *   the GNU General Public License for more details.
> - *
> - *   You should have received a copy of the GNU General Public License
> - *   along with this program;  if not, write to the Free Software
> - *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
>   */
>  
> -/*
> - * NAME
> - *	setitimer01.c
> - *
> - * DESCRIPTION
> - *	setitimer01 - check that a resonable setitimer() call succeeds.
> +/*\
> + * [Description]
>   *
> - * ALGORITHM
> - *	loop if that option was specified
> - *	allocate needed space and set up needed values
> - *	issue the system call
> - *	check the errno value
> - *	  issue a PASS message if we get zero
> - *	otherwise, the tests fails
> - *	  issue a FAIL message
> - *	  break any remaining tests
> - *	  call cleanup
> - *
> - * USAGE:  <for command-line>
> - *  setitimer01 [-c n] [-f] [-i n] [-I x] [-P x] [-t]
> - *     where,  -c n : Run n copies concurrently.
> - *             -f   : Turn off functionality Testing.
> - *	       -i n : Execute test n times.
> - *	       -I x : Execute test for x seconds.
> - *	       -P x : Pause for x seconds between iterations.
> - *	       -t   : Turn on syscall timing.
> - *
> - * HISTORY
> - *	03/2001 - Written by Wayne Boyer
> - *
> - * RESTRICTIONS
> - *	none
> + * Check that a setitimer() call pass with timer seting.
> + * Check if signal is generated correctly when when timer expiration.
>   */
>  
> -#include "test.h"
> -
>  #include <errno.h>
>  #include <sys/time.h>
> -
> -void cleanup(void);
> -void setup(void);
> -
> -char *TCID = "setitimer01";
> -int TST_TOTAL = 1;
> +#include <stdlib.h>
> +#include "tst_test.h"
> +#include "lapi/syscalls.h"
>  
>  #define SEC0	0
> -#define SEC1	20
> -#define SEC2	40
> -
> -int main(int ac, char **av)
> +#define SEC1	3
> +#define SEC2	5
> +
> +static volatile int si_flag;
> +static struct itimerval *value, *ovalue;
> +
> +static struct tcase {
> +       int which;
> +       struct itimerval **val;
> +       struct itimerval **oval;
> +       int signo;
> +} tcases[] = {
> +       {ITIMER_REAL,    &value, &ovalue, SIGALRM},
> +       {ITIMER_VIRTUAL, &value, &ovalue, SIGVTALRM},
> +       {ITIMER_PROF,    &value, &ovalue, SIGPROF},

Same issue here as in other itimer test.

> +};
> +
> +static int sys_setitimer(int which, void *new_value, void *old_value)
>  {
> -	int lc;
> -	struct itimerval *value, *ovalue;
> -
> -	tst_parse_opts(ac, av, NULL, NULL);
> -
> -	setup();		/* global setup */
> -
> -	/* The following loop checks looping state if -i option given */
> -
> -	for (lc = 0; TEST_LOOPING(lc); lc++) {
> -		/* reset tst_count in case we are looping */
> -		tst_count = 0;
> -
> -		/* allocate some space for the timer structures */
> -
> -		if ((value = malloc((size_t)sizeof(struct itimerval))) ==
> -		    NULL) {
> -			tst_brkm(TBROK, cleanup, "value malloc failed");
> -		}
> -
> -		if ((ovalue = malloc((size_t)sizeof(struct itimerval))) ==
> -		    NULL) {
> -			tst_brkm(TBROK, cleanup, "ovalue malloc failed");
> -		}
> -
> -		/* set up some reasonable values */
> -
> -		value->it_value.tv_sec = SEC1;
> -		value->it_value.tv_usec = SEC0;
> -		value->it_interval.tv_sec = 0;
> -		value->it_interval.tv_usec = 0;
> -		/*
> -		 * issue the system call with the TEST() macro
> -		 * ITIMER_REAL = 0, ITIMER_VIRTUAL = 1 and ITIMER_PROF = 2
> -		 */
> +	return tst_syscall(__NR_setitimer, which, new_value, old_value);
> +}
>  
> -		TEST(setitimer(ITIMER_REAL, value, ovalue));
> +static void sig_routine(int signo)
> +{
> +	switch(signo){
> +	case SIGALRM:
> +		si_flag = 1;
> +		break;
> +	case SIGVTALRM:
> +		si_flag = 2;
> +		break;
> +	case SIGPROF:
> +		si_flag = 3;
> +		break;
> +	default:
> +		break;
> +	}
> +}
>  
> -		if (TEST_RETURN != 0) {
> -			tst_resm(TFAIL, "call failed - errno = %d - %s",
> -				 TEST_ERRNO, strerror(TEST_ERRNO));
> -			continue;
> +static void verify_setitimer(unsigned int i)
> +{
> +	struct tcase *tc = &tcases[i];
> +
> +	si_flag = 0;
> +	value->it_value.tv_sec = SEC1;
> +	value->it_value.tv_usec = SEC0;

Why not test usecs instead?

> +	value->it_interval.tv_sec = SEC0;
> +	value->it_interval.tv_usec = SEC0;
> +
> +	TST_EXP_PASS(sys_setitimer(tc->which, *(tc->val), *(tc->oval)));
> +
> +	/*
> +	 * call setitimer again with new values.
> +	 * the old values should be stored in ovalue
> +	 */

We should probably set ovalue to some value > SEC1 here. Otherwise it
could just be zero or some previous value that is still valid.

> +	value->it_value.tv_sec = SEC2;
> +	value->it_value.tv_usec = SEC0;
> +
> +	TST_EXP_PASS(sys_setitimer(tc->which, *(tc->val), *(tc->oval)));
> +
> +	if (ovalue->it_value.tv_sec <= SEC1)
> +		tst_res(TPASS, "setitimer functionality is correct");
> +	else
> +		tst_brk(TFAIL, "old timer value is not equal to expected value");
> +
> +	for (;;) {

Could we use sigwait here instead?

We seem to be burning CPU cycles for no reason and if we get a spurious
signal the test will still pass if we get the correct one afterwards.

-- 
Thank you,
Richard.

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