Re: [dpdk-dev] [PATCH v6 1/3] ethdev: support device reset and recovery events

[Andrew Rybchenko <Andrew.Rybchenko@oktetlabs.ru>]

On 10/12/20 12:29 AM, Thomas Monjalon wrote:
> 09/10/2020 05:48, Kalesh A P:
>> From: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
>>
>> Adding support for device reset and recovery events in the
>> rte_eth_event framework. FW error and FW reset conditions would be
>> managed internally by PMD without needing application intervention.
>> In such cases, PMD would need reset/recovery events to notify application
>> that PMD is undergoing a reset.
>>
>> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
>> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
>> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
>> Reviewed-by: Asaf Penso <asafp@nvidia.com>
> 
> The ethdev maintainers are not Cc'ed.
> Please use the option --cc-cmd devtools/get-maintainer.sh
> 
> 
>> +Error recovery support
>> +~~~~~~~~~~~~~~~~~~~~~~
>> +
>> +When the PMD detects a FW reset or error condition, it will try to recover
>> +from the error without needing the application intervention. In such cases,
>> +PMD would need events to notify the application that it is undergoing
>> +an error recovery.
>> +
>> +The PMD will trigger RTE_ETH_EVENT_ERR_RECOVERING event to notify the
>> +application that PMD detected a FW reset or FW error condition. PMD will
>> +try to recover from the error by itself. Data path will be halted and
>> +control path operations would fail during the recovery period.
>> +
>> +The PMD will trigger RTE_ETH_EVENT_RECOVERED event to notify the application
>> +that the it has recovered from the error condition. Control path and data path
>> +are up now. Since the device undergone a reset, flow rules offloaded prior to
>> +the reset will be lost and the application has to recreate the rules again.

What should be done if the state is not recoverable?

>> diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h
>> index 9759f13..9b4b015 100644
>> --- a/lib/librte_ethdev/rte_ethdev.h
>> +++ b/lib/librte_ethdev/rte_ethdev.h
>> @@ -3207,6 +3207,23 @@ enum rte_eth_event_type {
>>  	RTE_ETH_EVENT_DESTROY,  /**< port is released */
>>  	RTE_ETH_EVENT_IPSEC,    /**< IPsec offload related event */
>>  	RTE_ETH_EVENT_FLOW_AGED,/**< New aged-out flows is detected */
>> +	RTE_ETH_EVENT_ERR_RECOVERING,
>> +			/**< port recovering from an error
>> +			 *
>> +			 * PMD detected a FW reset or error condition.
>> +			 * PMD will try to recover from the error.
>> +			 * Data path will be halted and Control path operations
>> +			 * would fail at this time.
>> +			 */
> 
> Does it mean the application has nothing to do when receiving this event?
> I think the app should stop polling at least.
> 
>> +	RTE_ETH_EVENT_RECOVERED,
>> +			/**< port recovered from an error
>> +			 *
>> +			 * PMD has recovered from the error condition.
>> +			 * Control path and Data path are up now.
>> +			 * Since the device undergone a reset, flow rules
>> +			 * offloaded prior to the reset will be lost and
>> +			 * the application has to recreate the rules again.
>> +			 */
> 
> Please be more precise.
> Should the app re-configure the port, setup the queues, start the port?
> 
>