public inbox for ltp@lists.linux.it
 help / color / mirror / Atom feed
* [LTP] [PATCH] msgstress04: make sure not pass IPC_PRIVATE to msgget
@ 2018-11-08  9:33 kai.kang
  2018-11-09 10:24 ` Jan Stancek
  0 siblings, 1 reply; 3+ messages in thread
From: kai.kang @ 2018-11-08  9:33 UTC (permalink / raw)
  To: ltp

From: Kai Kang <kai.kang@windriver.com>

It fails to run case msgstress04:

$ ./msgstress04 -n 1 -c 1 -l 1

It only inits the first MSGMNI members of keyarray. So if argument off
of function dotest_iteration is greater than MSGMNI, the argument key
passed to function dotest() will be IPC_PRIVATE(0) which will finally
pass to syscall msgget(), then cause case fails.

Make sure get the value for key from the initialized members of keyarray.

Signed-off-by: Kai Kang <kai.kang@windriver.com>
---
 testcases/kernel/syscalls/ipc/msgstress/msgstress04.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
index 18fad4fbd..a357ce59a 100644
--- a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
+++ b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
@@ -198,7 +198,7 @@ static void dotest_iteration(int off)
 	memset(pidarray, 0, sizeof(pidarray));
 
 	for (i = 0; i < nprocs; i++) {
-		key = keyarray[off + i];
+		key = keyarray[(off + i) % MSGMNI];
 
 		if ((pid = FORK_OR_VFORK()) < 0)
 			tst_brkm(TFAIL, cleanup,
-- 
2.17.0


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* [LTP] [PATCH] msgstress04: make sure not pass IPC_PRIVATE to msgget
  2018-11-08  9:33 [LTP] [PATCH] msgstress04: make sure not pass IPC_PRIVATE to msgget kai.kang
@ 2018-11-09 10:24 ` Jan Stancek
  2018-11-12  7:16   ` Kang Kai
  0 siblings, 1 reply; 3+ messages in thread
From: Jan Stancek @ 2018-11-09 10:24 UTC (permalink / raw)
  To: ltp



----- Original Message -----
> From: Kai Kang <kai.kang@windriver.com>
> 
> It fails to run case msgstress04:
> 
> $ ./msgstress04 -n 1 -c 1 -l 1
> 
> It only inits the first MSGMNI members of keyarray. So if argument off
> of function dotest_iteration is greater than MSGMNI, the argument key
> passed to function dotest() will be IPC_PRIVATE(0) which will finally
> pass to syscall msgget(), then cause case fails.
> 
> Make sure get the value for key from the initialized members of keyarray.
> 
> Signed-off-by: Kai Kang <kai.kang@windriver.com>
> ---
>  testcases/kernel/syscalls/ipc/msgstress/msgstress04.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
> b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
> index 18fad4fbd..a357ce59a 100644
> --- a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
> +++ b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
> @@ -198,7 +198,7 @@ static void dotest_iteration(int off)
>  	memset(pidarray, 0, sizeof(pidarray));
>  
>  	for (i = 0; i < nprocs; i++) {
> -		key = keyarray[off + i];
> +		key = keyarray[(off + i) % MSGMNI];

Hi,

Isn't this going to pick same key over and over again?

Say maxnprocs is '1', and MSGMNI is 10000:
  dotest_iteration(i*(10000 / 1));
will become:
  key = keyarray[(10000 + 0) % 10000];
  key = keyarray[(20000 + 0) % 10000];
  key = keyarray[(30000 + 0) % 10000];
  ...

If test intention is to exercise queues for all keys, shouldn't we fix main() instead?

diff --git a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
index 18fad4fbdca1..81383d6c8b11 100644
--- a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
+++ b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
@@ -176,11 +176,11 @@ int main(int argc, char **argv)
        } else {
                for (i = 0; i < (MSGMNI / maxnprocs); i++) {
                        nprocs = maxnprocs;
-                       dotest_iteration(i*(MSGMNI / maxnprocs));
+                       dotest_iteration(i * maxnprocs);
                }
 
                nprocs = MSGMNI % maxnprocs;
-               dotest_iteration(i*(MSGMNI / maxnprocs));
+               dotest_iteration(i * maxnprocs);
        }
 
        tst_resm(TPASS, "Test ran successfully!");

Regards,
Jan

>  
>  		if ((pid = FORK_OR_VFORK()) < 0)
>  			tst_brkm(TFAIL, cleanup,
> --
> 2.17.0
> 
> 
> --
> Mailing list info: https://lists.linux.it/listinfo/ltp
> 

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* [LTP] [PATCH] msgstress04: make sure not pass IPC_PRIVATE to msgget
  2018-11-09 10:24 ` Jan Stancek
@ 2018-11-12  7:16   ` Kang Kai
  0 siblings, 0 replies; 3+ messages in thread
From: Kang Kai @ 2018-11-12  7:16 UTC (permalink / raw)
  To: ltp

On 2018/11/9 下午6:24, Jan Stancek wrote:
>
> ----- Original Message -----
>> From: Kai Kang <kai.kang@windriver.com>
>>
>> It fails to run case msgstress04:
>>
>> $ ./msgstress04 -n 1 -c 1 -l 1
>>
>> It only inits the first MSGMNI members of keyarray. So if argument off
>> of function dotest_iteration is greater than MSGMNI, the argument key
>> passed to function dotest() will be IPC_PRIVATE(0) which will finally
>> pass to syscall msgget(), then cause case fails.
>>
>> Make sure get the value for key from the initialized members of keyarray.
>>
>> Signed-off-by: Kai Kang <kai.kang@windriver.com>
>> ---
>>   testcases/kernel/syscalls/ipc/msgstress/msgstress04.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
>> b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
>> index 18fad4fbd..a357ce59a 100644
>> --- a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
>> +++ b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
>> @@ -198,7 +198,7 @@ static void dotest_iteration(int off)
>>   	memset(pidarray, 0, sizeof(pidarray));
>>   
>>   	for (i = 0; i < nprocs; i++) {
>> -		key = keyarray[off + i];
>> +		key = keyarray[(off + i) % MSGMNI];
> Hi,
>
> Isn't this going to pick same key over and over again?
>
> Say maxnprocs is '1', and MSGMNI is 10000:
>    dotest_iteration(i*(10000 / 1));
> will become:
>    key = keyarray[(10000 + 0) % 10000];
>    key = keyarray[(20000 + 0) % 10000];
>    key = keyarray[(30000 + 0) % 10000];
>    ...
>
> If test intention is to exercise queues for all keys, shouldn't we fix main() instead?


Yes, you're right. Thanks for comment. V2 will be sent.

--Kai


>
> diff --git a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
> index 18fad4fbdca1..81383d6c8b11 100644
> --- a/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
> +++ b/testcases/kernel/syscalls/ipc/msgstress/msgstress04.c
> @@ -176,11 +176,11 @@ int main(int argc, char **argv)
>          } else {
>                  for (i = 0; i < (MSGMNI / maxnprocs); i++) {
>                          nprocs = maxnprocs;
> -                       dotest_iteration(i*(MSGMNI / maxnprocs));
> +                       dotest_iteration(i * maxnprocs);
>                  }
>   
>                  nprocs = MSGMNI % maxnprocs;
> -               dotest_iteration(i*(MSGMNI / maxnprocs));
> +               dotest_iteration(i * maxnprocs);
>          }
>   
>          tst_resm(TPASS, "Test ran successfully!");
>
> Regards,
> Jan
>
>>   
>>   		if ((pid = FORK_OR_VFORK()) < 0)
>>   			tst_brkm(TFAIL, cleanup,
>> --
>> 2.17.0
>>
>>
>> --
>> Mailing list info: https://lists.linux.it/listinfo/ltp
>>

-- 
Regards,
Neil | Kai Kang


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2018-11-12  7:16 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-11-08  9:33 [LTP] [PATCH] msgstress04: make sure not pass IPC_PRIVATE to msgget kai.kang
2018-11-09 10:24 ` Jan Stancek
2018-11-12  7:16   ` Kang Kai

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox