[PATCH v1 0/1] baseband/acc: fix queue setup failure clean up

[PATCH v1 0/1] baseband/acc: fix queue setup failure clean up

[Nicolas Chautru]

Hi Maxime, 

This is the fix for the regression issue mentioned off list.
This could notably hit when running with bbdev-test over
more than 64 threads.

Thanks
Nic


Nicolas Chautru (1):
  baseband/acc: fix queue setup failure clean up

 drivers/baseband/acc/rte_vrb_pmd.c | 2 ++
 1 file changed, 2 insertions(+)

-- 
2.34.1

[PATCH v1 1/1] baseband/acc: fix queue setup failure clean up

[Nicolas Chautru]

Incomplete clean up for vrb2 variant when queue setup fails
to complete successfully.

Fixes: fc65d3dcabe01 ("baseband/acc: refactor queue allocation")
Cc: stable@dpdk.org

Signed-off-by: Nicolas Chautru <nicolas.chautru@intel.com>
---
 drivers/baseband/acc/rte_vrb_pmd.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/baseband/acc/rte_vrb_pmd.c b/drivers/baseband/acc/rte_vrb_pmd.c
index 27620ccc10..2c10ef2c94 100644
--- a/drivers/baseband/acc/rte_vrb_pmd.c
+++ b/drivers/baseband/acc/rte_vrb_pmd.c
@@ -1156,6 +1156,8 @@ vrb_queue_setup(struct rte_bbdev *dev, uint16_t queue_id,
 free_q:
 	rte_free(q);
 	q = NULL;
+	if (d->device_variant == VRB2_VARIANT)
+		d->queue_index[conf->op_type]--;
 
 	return ret;
 }
-- 
2.34.1

Re: [PATCH v1 1/1] baseband/acc: fix queue setup failure clean up

[Patrick Robb]

[-- Attachment #1: Type: text/plain, Size: 166 bytes --]

Recheck-request: iol-intel-Performance

There was an infra failure with the Arm Grace server about 12 hours ago
which caused this failure - sending a retest request.

[-- Attachment #2: Type: text/html, Size: 204 bytes --]

Re: [PATCH v1 1/1] baseband/acc: fix queue setup failure clean up

[Maxime Coquelin]

On 3/5/25 7:27 PM, Nicolas Chautru wrote:
> Incomplete clean up for vrb2 variant when queue setup fails
> to complete successfully.
> 
> Fixes: fc65d3dcabe01 ("baseband/acc: refactor queue allocation")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Nicolas Chautru <nicolas.chautru@intel.com>
> ---
>   drivers/baseband/acc/rte_vrb_pmd.c | 2 ++
>   1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/baseband/acc/rte_vrb_pmd.c b/drivers/baseband/acc/rte_vrb_pmd.c
> index 27620ccc10..2c10ef2c94 100644
> --- a/drivers/baseband/acc/rte_vrb_pmd.c
> +++ b/drivers/baseband/acc/rte_vrb_pmd.c
> @@ -1156,6 +1156,8 @@ vrb_queue_setup(struct rte_bbdev *dev, uint16_t queue_id,
>   free_q:
>   	rte_free(q);
>   	q = NULL;
> +	if (d->device_variant == VRB2_VARIANT)
> +		d->queue_index[conf->op_type]--;
>   
>   	return ret;
>   }

Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>

Thanks,
Maxime

Re: [PATCH v1 1/1] baseband/acc: fix queue setup failure clean up

[Maxime Coquelin]

On 3/12/25 2:48 PM, Maxime Coquelin wrote:
> 
> 
> On 3/5/25 7:27 PM, Nicolas Chautru wrote:
>> Incomplete clean up for vrb2 variant when queue setup fails
>> to complete successfully.
>>
>> Fixes: fc65d3dcabe01 ("baseband/acc: refactor queue allocation")
>> Cc: stable@dpdk.org
>>
>> Signed-off-by: Nicolas Chautru <nicolas.chautru@intel.com>
>> ---
>>   drivers/baseband/acc/rte_vrb_pmd.c | 2 ++
>>   1 file changed, 2 insertions(+)
>>
>> diff --git a/drivers/baseband/acc/rte_vrb_pmd.c b/drivers/baseband/ 
>> acc/rte_vrb_pmd.c
>> index 27620ccc10..2c10ef2c94 100644
>> --- a/drivers/baseband/acc/rte_vrb_pmd.c
>> +++ b/drivers/baseband/acc/rte_vrb_pmd.c
>> @@ -1156,6 +1156,8 @@ vrb_queue_setup(struct rte_bbdev *dev, uint16_t 
>> queue_id,
>>   free_q:
>>       rte_free(q);
>>       q = NULL;
>> +    if (d->device_variant == VRB2_VARIANT)
>> +        d->queue_index[conf->op_type]--;
>>       return ret;
>>   }
> 
> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
> 
> Thanks,
> Maxime
> 

Applied to next-baseband.

Thanks,
Maxime