Netdev Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Jussi Kivilinna <jussi.kivilinna@haltian.com>
To: Antonio Quartulli <a@unstable.cc>,
	The list for a Better Approach To Mobile Ad-hoc Networking
	<b.a.t.m.a.n@lists.open-mesh.org>,
	netdev@vger.kernel.org
Subject: Is netif_rx_ni safe from interrupt context? (Re: [PATCH] batman-adv: bla: use netif_rx_ni when not in interrupt context)
Date: Wed, 19 Aug 2020 10:39:52 +0300	[thread overview]
Message-ID: <92c88618-23ec-2e30-1d79-b84f8d193bc3@haltian.com> (raw)
In-Reply-To: <987760ed-17d2-d486-5dd1-d40418f62ac1@unstable.cc>

Hello,

+CC netdev mailing-list

On 18.8.2020 23.12, Antonio Quartulli wrote:
> Hi,
> 
> On 18/08/2020 16:46, Jussi Kivilinna wrote:
>> batadv_bla_send_claim() gets called from worker thread context through
>> batadv_bla_periodic_work(), thus netif_rx_ni needs to be used in that
>> case. This fixes "NOHZ: local_softirq_pending 08" log messages seen
>> when batman-adv is enabled.
>>
>> Signed-off-by: Jussi Kivilinna <jussi.kivilinna@haltian.com>
>> ---
>>  net/batman-adv/bridge_loop_avoidance.c | 5 ++++-
>>  1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/net/batman-adv/bridge_loop_avoidance.c b/net/batman-adv/bridge_loop_avoidance.c
>> index 5c41cc52bc53..ab6cec3c7586 100644
>> --- a/net/batman-adv/bridge_loop_avoidance.c
>> +++ b/net/batman-adv/bridge_loop_avoidance.c
>> @@ -437,7 +437,10 @@ static void batadv_bla_send_claim(struct batadv_priv *bat_priv, u8 *mac,
>>  	batadv_add_counter(bat_priv, BATADV_CNT_RX_BYTES,
>>  			   skb->len + ETH_HLEN);
>>  
>> -	netif_rx(skb);
>> +	if (in_interrupt())
>> +		netif_rx(skb);
>> +	else
>> +		netif_rx_ni(skb);
> 
> What's the downside in calling netif_rx_ni() all the times?
> Is there any possible side effect?
> (consider this call is not along the fast path)

Good question. I tried to find answer for this but found documentation being lacking 
on the issue, so I looked for examples and used 'in_interrupt/netif_rx/netif_rx_ni' 
bit that appears in few other places: 
 https://elixir.bootlin.com/linux/v5.8/source/drivers/net/caif/caif_hsi.c#L469
 https://elixir.bootlin.com/linux/v5.8/source/drivers/net/wireless/broadcom/brcm80211/brcmfmac/core.c#L425
 https://elixir.bootlin.com/linux/v5.8/source/drivers/net/wireless/marvell/libertas/rx.c#L153
 https://elixir.bootlin.com/linux/v5.8/source/drivers/net/wireless/marvell/mwifiex/uap_txrx.c#L356
 https://elixir.bootlin.com/linux/v5.8/source/net/caif/chnl_net.c#L121

Maybe someone on netdev mailing-list could give hint on this matter - should 
'in_interrupt()?netif_rx(skb):netif_rx_ni(skb)' be used if context is not known or 
is just using 'netif_rx_ni(skb)' ok? 

> 
> On top of that, I just checked the definition of in_interrupt() and I
> got this comment:
> 
>  * Note: due to the BH disabled confusion: in_softirq(),in_interrupt()
> really
>  *       should not be used in new code.
> 
> 
> Check
> https://elixir.bootlin.com/linux/latest/source/include/linux/preempt.h#L85
> 
> Is that something we should consider or is the comment bogus?

It very well be that the existing code that I looked at may not be the best 
for reuse today.

-Jussi

           reply	other threads:[~2020-08-19  7:40 UTC|newest]

Thread overview: expand[flat|nested]  mbox.gz  Atom feed
 [parent not found: <987760ed-17d2-d486-5dd1-d40418f62ac1@unstable.cc>]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=92c88618-23ec-2e30-1d79-b84f8d193bc3@haltian.com \
    --to=jussi.kivilinna@haltian.com \
    --cc=a@unstable.cc \
    --cc=b.a.t.m.a.n@lists.open-mesh.org \
    --cc=netdev@vger.kernel.org \
    --subject='Is netif_rx_ni safe from interrupt context? (Re: [PATCH] batman-adv: bla: use netif_rx_ni when not in interrupt context)' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).