Broadcasting…
Regards,
Hiral
On 7/20/15, 5:48 PM, "Hiral Shah (hishah)" <hishah@xxxxxxxxx> wrote:
>Hi Martin,
>
>Thanks for the suggestion. Actually boot make more sense as the value will be either 0 or 1 only. We have already release following patch to other non-linux customers. We will make sure next time.
>
>
>Regards,
>Hiral
>
>
>
>
>
>
>
>
>On 7/16/15, 9:56 PM, "Martin K. Petersen" <martin.petersen@xxxxxxxxxx> wrote:
>
>>>>>>> "Hiral" == Hiral Shah <hishah@xxxxxxxxx> writes:
>>
>>Hiral> We added changes in fnic driver patch 1.6.0.16 to acquire
>>Hiral> io_req_lock in fnic_queuecommand() before issuing I/O so that io
>>Hiral> completion is serialized. But when releasing the lock we check
>>Hiral> for the I/O flag and this could be modified if IO abort occurs
>>Hiral> before I/O completion. In this case we wont release the lock and
>>Hiral> causes deadlock in some scenerios. Using the local variable to
>>Hiral> check the IO lock status will resolve the problem.
>>
>>Maybe bool instead of int?
>>
>>Otherwise OK.
>>
>>Reviewed-by: Martin K. Petersen <martin.petersen@xxxxxxxxxx>
>>
>>--
>>Martin K. Petersen Oracle Linux Engineering
��.n��������+%������w��{.n�����{������ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
