[Accel-config] Re: [PATCH v1 1/1] accel-config/test: Fix description timeout calculation to overflow
Zhu, Tony <tony.zhu at intel.com> Thu, 28 Apr 2022 02:31:04 +0000
| Newsgroups | dev.linux.lists.accel-config |
|---|---|
| Message-ID | <BN9PR11MB543326A382FD451895B557168AFD9@BN9PR11MB5433.namprd11.prod.outlook.com> |
It is used in other functions, could not be removed.
Tony(zhu, xinzhan)
Cube:SHZ1-3W-279
iNet:8821-5077
-----Original Message-----
From: Jiang, Dave <dave.jiang(a)intel.com>
Sent: Thursday, April 28, 2022 10:20 AM
To: Zhu, Tony <tony.zhu(a)intel.com>; accel-config(a)lists.01.org
Cc: Thomas, Ramesh <ramesh.thomas(a)intel.com>; Buchanan, Kenneth <kenneth.buchanan(a)intel.com>
Subject: Re: [PATCH v1 1/1] accel-config/test: Fix description timeout calculation to overflow
On 4/27/2022 7:15 PM, Zhu, Tony wrote:
> Dave,
>
> The code base is pending branch. It included the common code refactoring.
>
> I've tested it. msec_timeout is defined in the function.
>
> int acctest_wait_on_desc_timeout(struct completion_record *comp,
> struct acctest_context *ctx,
> unsigned int msec_timeout)
>
> unsigned int ms_timeout = 5000, should not use it in this function.
Gotcha. Should the unused var be removed? Otherwise Reviewed-by: Dave Jiang <dave.jiang(a)intel.com>
>
> Tony(zhu, xinzhan)
> Cube:SHZ1-3W-279
> iNet:8821-5077
>
> -----Original Message-----
> From: Jiang, Dave <dave.jiang(a)intel.com>
> Sent: Thursday, April 28, 2022 9:49 AM
> To: Zhu, Tony <tony.zhu(a)intel.com>; accel-config(a)lists.01.org
> Cc: Thomas, Ramesh <ramesh.thomas(a)intel.com>; Buchanan, Kenneth
> <kenneth.buchanan(a)intel.com>
> Subject: Re: [PATCH v1 1/1] accel-config/test: Fix description timeout
> calculation to overflow
>
>
> On 4/27/2022 6:31 PM, Tony Zhu wrote:
>> Dsa_test application will fail when when utilizing memory sizes
>> larger than 128MiB against multiple devices. The init of timeout
>> should use the input msec_timeout.
>>
>> Signed-off-by: Tony Zhu <tony.zhu(a)intel.com>
> Ken, you want your sign off added?
>
>> ---
>> test/accel_test.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/test/accel_test.c b/test/accel_test.c index
>> 8487099..0468a00 100644
>> --- a/test/accel_test.c
>> +++ b/test/accel_test.c
>> @@ -331,7 +331,7 @@ int acctest_wait_on_desc_timeout(struct completion_record *comp,
>> j++;
>> }
>> } else {
>> - unsigned long timeout = (ms_timeout * 1000000) * 3;
>> + unsigned long timeout = (msec_timeout * 1000000ul) * 3;
> Couple things. This looks like changed with the common code refactoring?
> Has that merged by Ramesh? If not then maybe do the diff before that code since we need to get this fix released. And then make sure the refactoring take this change into account.
>
> Also, did you compile test this? I don't see the var name ms_timeout changed to msec_timeout.
>
>> int r = 1;
>> unsigned long t = 0;
>>