Re: [PATCH v2 net-next 1/5] net: dsa: microchip: add rmon grouping for ethtool statistics

[<Rakesh.Sankaranarayanan@microchip.com>]

On Fri, 2023-02-17 at 15:59 +0100, Alexander Lobakin wrote:
> [Some people who received this message don't often get email from
> aleksander.lobakin@intel.com. Learn why this is important at
> https://aka.ms/LearnAboutSenderIdentification ]
> 
> EXTERNAL EMAIL: Do not click links or open attachments unless you
> know the content is safe
> 
> From: Rakesh Sankaranarayanan <rakesh.sankaranarayanan@microchip.com>
> Date: Fri, 17 Feb 2023 16:32:07 +0530
> 
> >     Add support for ethtool standard device statistics grouping.
> > Support rmon
> >     statistics grouping using rmon groups parameter in ethtool
> > command. rmon
> >     provides packet size based range grouping. Common mib
> > parameters are used
> >     across all KSZ series swtches for packet size statistics,
> > except for
> >     KSZ8830. KSZ series have mib counters for packets with size:
> 
> [...]
> 
> > +void ksz8_get_rmon_stats(struct ksz_device *dev, int port,
> > +                      struct ethtool_rmon_stats *rmon_stats,
> > +                      const struct ethtool_rmon_hist_range
> > **ranges)
> > +{
> > +     struct ksz_port_mib *mib;
> > +     u64 *cnt;
> 
> Nit: I guess it can be const since you only read it (in every such
> callback)?
This variable is getting updated on every call back after reading from
registers, so we cannot have it as constant.

> 
> > +     u8 i;
> > +
> > +     mib = &dev->ports[port].mib;
> > +
> > +     mutex_lock(&mib->cnt_mutex);
> > +
> > +     cnt = &mib->counters[KSZ8_RX_UNDERSIZE];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ8_RX_UNDERSIZE, NULL,
> > cnt);
> > +     rmon_stats->undersize_pkts = *cnt;
> > +
> > +     cnt = &mib->counters[KSZ8_RX_OVERSIZE];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ8_RX_OVERSIZE, NULL,
> > cnt);
> > +     rmon_stats->oversize_pkts = *cnt;
> > +
> > +     cnt = &mib->counters[KSZ8_RX_FRAGMENTS];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ8_RX_FRAGMENTS, NULL,
> > cnt);
> > +     rmon_stats->fragments = *cnt;
> > +
> > +     cnt = &mib->counters[KSZ8_RX_JABBERS];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ8_RX_JABBERS, NULL,
> > cnt);
> > +     rmon_stats->jabbers = *cnt;
> > +
> > +     for (i = 0; i < KSZ8_HIST_LEN; i++) {
> > +             cnt = &mib->counters[KSZ8_RX_64_OR_LESS + i];
> > +             dev->dev_ops->r_mib_pkt(dev, port,
> > +                             (KSZ8_RX_64_OR_LESS + i), NULL, cnt);
> 
> Weird linewrap. Please align the following lines with the opening
> brace
> of the first one, e.g.
> 
>                 dev->dev_ops->r_mib_pkt(dev, port,
>                                         KSZ8_RX_64_OR_LESS + i, NULL,
>                                         cnt);
> 
> BUT I don't see why you need those braces around `macro + i` and
> without
> them you can fit it into the previous line I believe.
Will update it in next revision
> 
> > +             rmon_stats->hist[i] = *cnt;
> > +     }
> > +
> > +     mutex_unlock(&mib->cnt_mutex);
> > +
> > +     *ranges = ksz_rmon_ranges;
> > +}
> > +
> > +void ksz9477_get_rmon_stats(struct ksz_device *dev, int port,
> > +                         struct ethtool_rmon_stats *rmon_stats,
> > +                         const struct ethtool_rmon_hist_range
> > **ranges)
> > +{
> > +     struct ksz_port_mib *mib;
> > +     u64 *cnt;
> > +     u8 i;
> > +
> > +     mib = &dev->ports[port].mib;
> > +
> > +     mutex_lock(&mib->cnt_mutex);
> > +
> > +     cnt = &mib->counters[KSZ9477_RX_UNDERSIZE];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ9477_RX_UNDERSIZE,
> > NULL, cnt);
> > +     rmon_stats->undersize_pkts = *cnt;
> > +
> > +     cnt = &mib->counters[KSZ9477_RX_OVERSIZE];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ9477_RX_OVERSIZE, NULL,
> > cnt);
> > +     rmon_stats->oversize_pkts = *cnt;
> > +
> > +     cnt = &mib->counters[KSZ9477_RX_FRAGMENTS];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ9477_RX_FRAGMENTS,
> > NULL, cnt);
> > +     rmon_stats->fragments = *cnt;
> > +
> > +     cnt = &mib->counters[KSZ9477_RX_JABBERS];
> > +     dev->dev_ops->r_mib_pkt(dev, port, KSZ9477_RX_JABBERS, NULL,
> > cnt);
> > +     rmon_stats->jabbers = *cnt;
> > +
> > +     for (i = 0; i < KSZ9477_HIST_LEN; i++) {
> > +             cnt = &mib->counters[KSZ9477_RX_64_OR_LESS + i];
> > +             dev->dev_ops->r_mib_pkt(dev, port,
> > +                             (KSZ9477_RX_64_OR_LESS + i), NULL,
> > cnt);
> 
> (same, and please check all other places in the series)
sure, will update in next version.
> 
> > +             rmon_stats->hist[i] = *cnt;
> > +     }
> > +
> > +     mutex_unlock(&mib->cnt_mutex);
> > +
> > +     *ranges = ksz_rmon_ranges;
> > +}
> [...]
> 
> Thanks,
> Olek