[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200803151818.5a2e5616@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
Date: Mon, 3 Aug 2020 15:18:18 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: <sameehj@...zon.com>
Cc: <davem@...emloft.net>, <netdev@...r.kernel.org>, <dwmw@...zon.com>,
<zorik@...zon.com>, <matua@...zon.com>, <saeedb@...zon.com>,
<msw@...zon.com>, <aliguori@...zon.com>, <nafea@...zon.com>,
<gtzalik@...zon.com>, <netanel@...zon.com>, <alisaidi@...zon.com>,
<benh@...zon.com>, <akiyano@...zon.com>, <ndagan@...zon.com>
Subject: Re: [PATCH V1 net-next 1/3] net: ena: ethtool: Add new device
statistics
On Sat, 1 Aug 2020 14:21:28 +0000 sameehj@...zon.com wrote:
> + if (eni_stats_needed) {
> + ena_update_hw_stats(adapter);
> + for (i = 0; i < ENA_STATS_ARRAY_ENI(adapter); i++) {
> + ena_stats = &ena_stats_eni_strings[i];
> +
> + ptr = (u64 *)((uintptr_t)&adapter->eni_stats +
> + (uintptr_t)ena_stats->stat_offset);
In the kernel unsigned long is the type for doing maths on pointers.
> + ena_safe_update_stat(ptr, data++, &adapter->syncp);
> + }
> + }
> +
> ena_queue_stats(adapter, &data);
> ena_dev_admin_queue_stats(adapter, &data);
> }
>
> +static void ena_get_ethtool_stats(struct net_device *netdev,
> + struct ethtool_stats *stats,
> + u64 *data)
> +{
> + struct ena_adapter *adapter = netdev_priv(netdev);
> +
> + ena_get_stats(adapter, data, adapter->eni_stats_supported);
> +}
Why the indirections? You always pass adapter->eni_stats_supported as a
parameter, why not just use it directly?
Other than the two nits, the set LGTM.
Powered by blists - more mailing lists