Hi zoran,

comments inline.

Thanks,
Neel.

On Monday 18 January 2016 06:46 PM, Zoran Milinkovic wrote:
> Hi Neelakanta,
>
> If you only change osafassert, then the similar osafassert must be added in 
> immnd_evt_proc_ccb_finalize() after immnd_evt_ccb_abort().
> And another check must be added in "do-while" loop in 
> immnd_evt_proc_ccb_finalize() for checking that clientArr[ix] is not 0.
This is not required,  as clientArr will be accessed only if clArrSize 
is more than 1 in immnd_evt_proc_ccb_finalize.
> I feel more safely if client connections with 0 are removed from client array.
>
> Best regards,
> Zoran
>
>
> -----Original Message-----
> From: Neelakanta Reddy [mailto:[email protected]]
> Sent: Monday, January 18, 2016 12:37 PM
> To: Zoran Milinkovic; Hung Duc Nguyen
> Cc: [email protected]
> Subject: Re: [PATCH 1 of 1] imm: corrected the assert checking in CCb apply 
> when CCB is aborted [#1664]
>
> Hi zoran,
>
> yes, clArr should be freed here also. I added in finalize, but forgot to add 
> here.
> Re publish the patch again.
> comments inline.
>
> Thanks,
> Neel.
> On Monday 18 January 2016 04:47 PM, Zoran Milinkovic wrote:
>> Hi Neelakanta,
>>
>> Memory of clArr is not freed. It was one of comments in the review of #1503.
>>
>> If client connection is 0, then it should be removed from clArr (clArrSize 
>> will be decreased), and osafassert can remain as it was before the patch.
>> The change can be done in immModel_ccbAbort() removing client connections of 
>> 0.
>>
>> In the patch, clArr can still point to NULL  if immModel_ccbAbort() in
>> immnd_evt_ccb_abort() returns true. Then osafassert will crash in try
>> to access clArr[0];
> The below modified assert can resolve when the clArr  NULL.
> osafassert(!clArrSize||!clArr[0] || originatedAtThisNd);
>> Nack from me.
>>
>> Best regards,
>> Zoran
>>
>>
>> -----Original Message-----
>> From: [email protected] [mailto:[email protected]]
>> Sent: Monday, January 18, 2016 7:28 AM
>> To: Zoran Milinkovic; Hung Duc Nguyen
>> Cc: [email protected]
>> Subject: [PATCH 1 of 1] imm: corrected the assert checking in CCb
>> apply when CCB is aborted [#1664]
>>
>>    osaf/services/saf/immsv/immnd/immnd_evt.c |  6 +++++-
>>    1 files changed, 5 insertions(+), 1 deletions(-)
>>
>>
>> diff --git a/osaf/services/saf/immsv/immnd/immnd_evt.c
>> b/osaf/services/saf/immsv/immnd/immnd_evt.c
>> --- a/osaf/services/saf/immsv/immnd/immnd_evt.c
>> +++ b/osaf/services/saf/immsv/immnd/immnd_evt.c
>> @@ -7899,7 +7899,11 @@ static void immnd_evt_proc_ccb_apply(IMM
>>                      /*err != SA_AIS_OK => generate SaImmOiCcbAbortCallbackT 
>> upcalls
>>                       */
>>                      immnd_evt_ccb_abort(cb, evt->info.ccbId, &clArr, 
>> &clArrSize, NULL);
>> -                    osafassert(!clArrSize || originatedAtThisNd);
>> +                    /* when the client is not originated from this ND then 
>> the client
>> +                       connection must be zero. We are in apply and there 
>> will be only
>> +                       one implementer connection and no augumentaion 
>> connections.
>> +                    */
>> +                    osafassert(!clArr[0] || originatedAtThisNd);
>>              }
>>              TRACE_2("CCB APPLY TERMINATING CCB: %u", evt->info.ccbId);
>>              bCcbFinalize = 1;


------------------------------------------------------------------------------
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=267308311&iu=/4140
_______________________________________________
Opensaf-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensaf-devel

Reply via email to