netdev - Re: [PATCH net-next v3 06/10] net/mlx5e: Change Mellanox references in DIM code

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 8 Jan 2018 23:09:00 -0800
From:   Saeed Mahameed <saeedm@...lanox.com>
To:     Andy Gospodarek <andy@...yhouse.net>, netdev@...r.kernel.org
Cc:     mchan@...adcom.com, talgi@...lanox.com, ogerlitz@...lanox.com,
        Andy Gospodarek <gospo@...adcom.com>
Subject: Re: [PATCH net-next v3 06/10] net/mlx5e: Change Mellanox references
 in DIM code



On 01/08/2018 11:06 PM, Saeed Mahameed wrote:
> 
> 
> On 01/08/2018 10:13 PM, Andy Gospodarek wrote:
>> From: Andy Gospodarek <gospo@...adcom.com>
>>
>> Change all appropriate mlx5_am* and MLX5_AM* references to net_dim and
>> NET_DIM, respectively, in code that handles dynamic interrupt
>> moderation.  Also change all references from 'am' to 'dim' when used as
>> local variables and add generic profile references.
>>
>> Signed-off-by: Andy Gospodarek <gospo@...adcom.com>
>> Acked-by: Tal Gilboa <talgi@...lanox.com>
>> Acked-by: Saeed Mahameed <saeedm@...lanox.com>
>> ---
>>   drivers/net/ethernet/mellanox/mlx5/core/en.h       |   9 +-
>>   drivers/net/ethernet/mellanox/mlx5/core/en_dim.c   |  14 +-
>>   .../net/ethernet/mellanox/mlx5/core/en_ethtool.c   |   6 +-
>>   drivers/net/ethernet/mellanox/mlx5/core/en_main.c  |  40 ++-
>>   drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c  |   8 +-
>>   drivers/net/ethernet/mellanox/mlx5/core/net_dim.c  | 286 
>> ++++++++++-----------
>>   drivers/net/ethernet/mellanox/mlx5/core/net_dim.h  |  63 ++---
>>   7 files changed, 225 insertions(+), 201 deletions(-)
>>
> 
> [...]
> 
>>   #define IS_SIGNIFICANT_DIFF(val, ref) \
>>       (((100 * abs((val) - (ref))) / (ref)) > 10) /* more than 10% 
>> difference */
>> -static int mlx5e_am_stats_compare(struct mlx5e_rx_am_stats *curr,
>> -                  struct mlx5e_rx_am_stats *prev)
>> +static int net_dim_stats_compare(struct net_dim_stats *curr,
>> +                 struct net_dim_stats *prev)
>>   {
>>       if (!prev->bpms)
>> -        return curr->bpms ? MLX5E_AM_STATS_BETTER :
>> -                    MLX5E_AM_STATS_SAME;
>> +        return curr->bpms ? NET_DIM_STATS_BETTER :
>> +                    NET_DIM_STATS_SAME;
>>       if (IS_SIGNIFICANT_DIFF(curr->bpms, prev->bpms))
>> -        return (curr->bpms > prev->bpms) ? MLX5E_AM_STATS_BETTER :
>> -                           MLX5E_AM_STATS_WORSE;
>> +        return (curr->bpms > prev->bpms) ? NET_DIM_STATS_BETTER :
>> +                           NET_DIM_STATS_WORSE;
> 
> Hey Andy,
> 
> I am currently reviewing a patch internally that fixes a bug in this 
> area, prev->ppms can be 0 and could cause IS_SIGNIFICANT_DIFF ouch !

I meant cause division by 0 in "IS_SIGNIFICANT_DIFF"

> same goes for prev->eppm, for some reason we had a broken assumption 
> that if ppms is 0 for some reason then the bpms is 0 and the above 
> condition will cover us.
> 
> Anyway the patch will go to net, which means when this series gets 
> accepted then net-next will fail to merge with net and we need to 
> manually push the fix to the new DIM library.
> 
> But for now I don't think anything is required for this series other 
> than bringing this division by 0 issue and the future merge conflict to 
> your attention.
> 
>>       if (IS_SIGNIFICANT_DIFF(curr->ppms, prev->ppms))
>> -        return (curr->ppms > prev->ppms) ? MLX5E_AM_STATS_BETTER :
>> -                           MLX5E_AM_STATS_WORSE;
>> +        return (curr->ppms > prev->ppms) ? NET_DIM_STATS_BETTER :
>> +                           NET_DIM_STATS_WORSE;
>>       if (IS_SIGNIFICANT_DIFF(curr->epms, prev->epms))
>> -        return (curr->epms < prev->epms) ? MLX5E_AM_STATS_BETTER :
>> -                           MLX5E_AM_STATS_WORSE;
>> +        return (curr->epms < prev->epms) ? NET_DIM_STATS_BETTER :
>> +                           NET_DIM_STATS_WORSE;
>> -    return MLX5E_AM_STATS_SAME;
>> +    return NET_DIM_STATS_SAME;
>>   }