idpf: Don't hard code napi_struct size

[net-next v3] idpf: Don't hard code napi_struct size

Posted by Joe Damato 1 year, 4 months ago

The sizeof(struct napi_struct) can change. Don't hardcode the size to
400 bytes and instead use "sizeof(struct napi_struct)".

Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Joe Damato <jdamato@fastly.com>
---
 drivers/net/ethernet/intel/idpf/idpf_txrx.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
index f0537826f840..9c1fe84108ed 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
+++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
@@ -438,7 +438,8 @@ struct idpf_q_vector {
 	__cacheline_group_end_aligned(cold);
 };
 libeth_cacheline_set_assert(struct idpf_q_vector, 112,
-			    424 + 2 * sizeof(struct dim),
+			    24 + sizeof(struct napi_struct) +
+			    2 * sizeof(struct dim),
 			    8 + sizeof(cpumask_var_t));
 
 struct idpf_rx_queue_stats {
-- 
2.34.1

Re: [Intel-wired-lan] [net-next v3] idpf: Don't hard code napi_struct size

Posted by Jacob Keller 9 months, 3 weeks ago


On 10/4/2024 3:54 AM, Joe Damato wrote:
> The sizeof(struct napi_struct) can change. Don't hardcode the size to
> 400 bytes and instead use "sizeof(struct napi_struct)".
> 
> Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
> Signed-off-by: Joe Damato <jdamato@fastly.com>
> ---
>  drivers/net/ethernet/intel/idpf/idpf_txrx.h | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> index f0537826f840..9c1fe84108ed 100644
> --- a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> +++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> @@ -438,7 +438,8 @@ struct idpf_q_vector {
>  	__cacheline_group_end_aligned(cold);
>  };
>  libeth_cacheline_set_assert(struct idpf_q_vector, 112,
> -			    424 + 2 * sizeof(struct dim),
> +			    24 + sizeof(struct napi_struct) +
> +			    2 * sizeof(struct dim),
>  			    8 + sizeof(cpumask_var_t));
>  
>  struct idpf_rx_queue_stats {

Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>

Re: [net-next v3] idpf: Don't hard code napi_struct size

Posted by Paolo Abeni 1 year, 4 months ago

On 10/4/24 12:54, Joe Damato wrote:
> The sizeof(struct napi_struct) can change. Don't hardcode the size to
> 400 bytes and instead use "sizeof(struct napi_struct)".
> 
> Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
> Signed-off-by: Joe Damato <jdamato@fastly.com>
> ---
>   drivers/net/ethernet/intel/idpf/idpf_txrx.h | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> index f0537826f840..9c1fe84108ed 100644
> --- a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> +++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> @@ -438,7 +438,8 @@ struct idpf_q_vector {
>   	__cacheline_group_end_aligned(cold);
>   };
>   libeth_cacheline_set_assert(struct idpf_q_vector, 112,
> -			    424 + 2 * sizeof(struct dim),
> +			    24 + sizeof(struct napi_struct) +
> +			    2 * sizeof(struct dim),
>   			    8 + sizeof(cpumask_var_t));
>   
>   struct idpf_rx_queue_stats {

@Tony: I'm assuming you want this one to go via your tree first, please 
LMK otherwise.

Thanks,

Paolo

Re: [Intel-wired-lan] [net-next v3] idpf: Don't hard code napi_struct size

Posted by Tony Nguyen 1 year, 4 months ago


On 10/8/2024 2:14 AM, Paolo Abeni wrote:
> On 10/4/24 12:54, Joe Damato wrote:
>> The sizeof(struct napi_struct) can change. Don't hardcode the size to
>> 400 bytes and instead use "sizeof(struct napi_struct)".
>>
>> Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
>> Signed-off-by: Joe Damato <jdamato@fastly.com>
>> ---
>>   drivers/net/ethernet/intel/idpf/idpf_txrx.h | 3 ++-
>>   1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h 
>> b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
>> index f0537826f840..9c1fe84108ed 100644
>> --- a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
>> +++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
>> @@ -438,7 +438,8 @@ struct idpf_q_vector {
>>       __cacheline_group_end_aligned(cold);
>>   };
>>   libeth_cacheline_set_assert(struct idpf_q_vector, 112,
>> -                424 + 2 * sizeof(struct dim),
>> +                24 + sizeof(struct napi_struct) +
>> +                2 * sizeof(struct dim),
>>                   8 + sizeof(cpumask_var_t));
>>   struct idpf_rx_queue_stats {
> 
> @Tony: I'm assuming you want this one to go via your tree first, please 
> LMK otherwise.

Hi Paolo,

This seems straightforward and we'd see the issues compile-time so I 
think it's ok for you to take directly.

Reviewed-by: Tony Nguyen <anthony.l.nguyen@intel.com>

Thanks,
Tony

Re: [net-next v3] idpf: Don't hard code napi_struct size

Posted by Alexander Lobakin 1 year, 4 months ago

From: Joe Damato <jdamato@fastly.com>
Date: Fri,  4 Oct 2024 10:54:07 +0000

> The sizeof(struct napi_struct) can change. Don't hardcode the size to
> 400 bytes and instead use "sizeof(struct napi_struct)".
> 
> Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
> Signed-off-by: Joe Damato <jdamato@fastly.com>

Acked-by: Alexander Lobakin <aleksander.lobakin@intel.com>

I'm sorry that I complicated things a bit with those assertions =\

> ---
>  drivers/net/ethernet/intel/idpf/idpf_txrx.h | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> index f0537826f840..9c1fe84108ed 100644
> --- a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> +++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
> @@ -438,7 +438,8 @@ struct idpf_q_vector {
>  	__cacheline_group_end_aligned(cold);
>  };
>  libeth_cacheline_set_assert(struct idpf_q_vector, 112,
> -			    424 + 2 * sizeof(struct dim),
> +			    24 + sizeof(struct napi_struct) +
> +			    2 * sizeof(struct dim),
>  			    8 + sizeof(cpumask_var_t));
>  
>  struct idpf_rx_queue_stats {

Thanks,
Olek