Re: [PATCH net-next 5/5] ionic: Add .get_fec_stats ethtool handler

[Vadim Fedorenko <vadim.fedorenko@linux.dev>]

On 05/05/2026 20:43, Eric Joyner wrote:
> On 5/5/2026 6:54 AM, Vadim Fedorenko wrote:
>> [You don't often get email from vadim.fedorenko@linux.dev. Learn why this is
>> important at https://aka.ms/LearnAboutSenderIdentification ]
>>
>> Caution: This message originated from an External Source. Use proper caution
>> when opening attachments, clicking links, or responding.
>>
>>
>> On 01/05/2026 04:15, Eric Joyner wrote:
>>> Several FEC error statistics being collected can be reported in a
>>> dedicated ethtool callback for FEC errors, so implement the handler that
>>> does so. This includes 802.3ck FEC histogram data that some newer
>>> hardware collects.
>>>
>>> Assisted-by: Claude:claude-4.6-sonnet
>>> Signed-off-by: Eric Joyner <eric.joyner@amd.com>
>>> ---
>>>    .../ethernet/pensando/ionic/ionic_ethtool.c   | 51 +++++++++++++++++++
>>>    1 file changed, 51 insertions(+)
>>>
>>> diff --git a/drivers/net/ethernet/pensando/ionic/ionic_ethtool.c b/drivers/
>>> net/ethernet/pensando/ionic/ionic_ethtool.c
>>> index 78a802eb159f..fe1f753b6115 100644
>>> --- a/drivers/net/ethernet/pensando/ionic/ionic_ethtool.c
>>> +++ b/drivers/net/ethernet/pensando/ionic/ionic_ethtool.c
>>> @@ -418,6 +418,56 @@ static int ionic_get_fecparam(struct net_device *netdev,
>>>        return 0;
>>>    }
>>>
>>> +static const struct ethtool_fec_hist_range ionic_fec_ranges[] = {
>>> +     { 0, 0},
>>> +     { 1, 1},
>>> +     { 2, 2},
>>> +     { 3, 3},
>>> +     { 4, 4},
>>> +     { 5, 5},
>>> +     { 6, 6},
>>> +     { 7, 7},
>>> +     { 8, 8},
>>> +     { 9, 9},
>>> +     { 10, 10},
>>> +     { 11, 11},
>>> +     { 12, 12},
>>> +     { 13, 13},
>>> +     { 14, 14},
>>> +     { 15, 15},
>>> +     { 0, 0},
>>> +};
>>> +
>>> +static void
>>> +ionic_fill_fec_hist(const struct ionic_port_extra_stats *extra_stats,
>>> +                 struct ethtool_fec_hist *hist)
>>> +{
>>> +     int i;
>>> +
>>> +     hist->ranges = ionic_fec_ranges;
>>> +     for (i = 0; i < ETHTOOL_FEC_HIST_MAX - 1; i++)
>>> +             hist->values[i].sum = extra_stats->fec_codeword_error_bin[i];
>>> +}
>>
>> ETHTOOL_FEC_HIST_MAX = 17, you defined 16 bins, but iterating over 15 of
>> them. Looks like bin {15, 15} will be lost in stats.
>>
>>
> 
> This looks correct to me -- (ETHTOOL_FEC_HIST_MAX - 1) = 16, so starting with
> i=0, it'll iterate through the 16 bins and ignore the 17th end marker bin. Bin
> 15 does get included and gets its sum set.

Ahh, yeah, you're right, smth was wrong with my math yesterday. btw, how 
does it work for different FEC? RS(528, 514) will not give you 16 bins..