Re: [RFC 2/5] net/ethtool: support get coalesce per queue

From: Ben Hutchings <ben@decadent.org.uk>
To: kan.liang@intel.com, netdev@vger.kernel.org, davem@davemloft.net,
	bwh@kernel.org
Cc: jesse.brandeburg@intel.com, andi@firstfloor.org,
	f.fainelli@gmail.com, alexander.duyck@gmail.com,
	jeffrey.t.kirsher@intel.com, shannon.nelson@intel.com,
	carolyn.wyborny@intel.com, donald.c.skidmore@intel.com,
	mitch.a.williams@intel.com, ogerlitz@mellanox.com,
	edumazet@google.com, jiri@mellanox.com, sfeldma@gmail.com,
	gospo@cumulusnetworks.com, sasha.levin@oracle.com,
	dsahern@gmail.com, tj@kernel.org, cascardo@redhat.com,
	corbet@lwn.net
Subject: Re: [RFC 2/5] net/ethtool: support get coalesce per queue
Date: Sat, 19 Dec 2015 03:35:43 +0000	[thread overview]
Message-ID: <1450496143.30687.10.camel@decadent.org.uk> (raw)
In-Reply-To: <1450335081-3012-2-git-send-email-kan.liang@intel.com>

[-- Attachment #1: Type: text/plain, Size: 2889 bytes --]

On Thu, 2015-12-17 at 06:51 +0000, kan.liang@intel.com wrote:
> From: Kan Liang <kan.liang@intel.com>
> 
> Device driver has to provide an interface to get per queue coalesce.
> The interrupt coalescing parameters of each masked queue will be
> copied back to user space one by one.
> 
> Signed-off-by: Kan Liang <kan.liang@intel.com>
> ---
>  include/linux/ethtool.h |  5 ++++-
>  net/core/ethtool.c      | 33 ++++++++++++++++++++++++++++++++-
>  2 files changed, 36 insertions(+), 2 deletions(-)
> 
> diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h
> index 653dc9c..107f75f 100644
> --- a/include/linux/ethtool.h
> +++ b/include/linux/ethtool.h
> @@ -201,6 +201,8 @@ static inline u32 ethtool_rxfh_indir_default(u32 index, u32 n_rx_rings)
>   * @get_module_eeprom: Get the eeprom information from the plug-in module
>   * @get_eee: Get Energy-Efficient (EEE) supported and status.
>   * @set_eee: Set EEE status (enable/disable) as well as LPI timers.
> + * @get_per_queue_coalesce: Get interrupt coalescing parameters per queue.
> + * 	Returns a negative error code or zero.

Is the ethtool core or the driver responsible for range-checking the
queue number?

What is reported if the queue number is in range for RX queues but not
for TX queues (or vice versa)?

>   *
>   * All operations are optional (i.e. the function pointer may be set
>   * to %NULL) and callers must take this into account.  Callers must
> @@ -279,7 +281,8 @@ struct ethtool_ops {
>  			       const struct ethtool_tunable *, void *);
>  	int	(*set_tunable)(struct net_device *,
>  			       const struct ethtool_tunable *, const void *);
> -
> +	int	(*get_per_queue_coalesce)(struct net_device *, int,
> +					  struct ethtool_coalesce *);
>  
>  };
>  #endif /* _LINUX_ETHTOOL_H */
> diff --git a/net/core/ethtool.c b/net/core/ethtool.c
> index 125fb32..22ff69a 100644
> --- a/net/core/ethtool.c
> +++ b/net/core/ethtool.c
> @@ -1748,6 +1748,36 @@ out:
>  	return ret;
>  }
>  
> +static int ethtool_get_per_queue_coalesce(struct net_device *dev,
> +					  void __user *useraddr,
> +					  struct ethtool_per_queue_op *per_queue_opt)
> +{
> +	u64 queue_mask;
> +	int bit, i, ret;
> +
> +	if (!dev->ethtool_ops->get_per_queue_coalesce)
> +		return -EOPNOTSUPP;
> +
> +	useraddr += sizeof(*per_queue_opt);
> +	for (i = 0; i < MAX_QUEUE_MASK; i++) {
> +		queue_mask = per_queue_opt->queue_mask[i];
> +		if (queue_mask > 0) {
> +			for_each_set_bit(bit, (unsigned long *)&queue_mask, 64) {
[...]

This assumes a little-endian system.

You should probably build on top of David Decotigny's recent patch
series that adds conversion of u32-based bitmaps to kernel bitmaps.

Ben.

-- 
Ben Hutchings
Always try to do things in chronological order;
it's less confusing that way.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 811 bytes --]