[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20181017.232458.1690499475868087571.davem@davemloft.net>
Date: Wed, 17 Oct 2018 23:24:58 -0700 (PDT)
From: David Miller <davem@...emloft.net>
To: hkallweit1@...il.com
Cc: romieu@...zoreil.com, nic_swsd@...ltek.com, netdev@...r.kernel.org
Subject: Re: [PATCH net] r8169: fix NAPI handling under high load
From: Heiner Kallweit <hkallweit1@...il.com>
Date: Thu, 18 Oct 2018 07:58:52 +0200
> On 18.10.2018 07:21, David Miller wrote:
>> From: Francois Romieu <romieu@...zoreil.com>
>> Date: Thu, 18 Oct 2018 01:30:45 +0200
>>
>>> Heiner Kallweit <hkallweit1@...il.com> :
>>> [...]
>>>> This issue has been there more or less forever (at least it exists in
>>>> 3.16 already), so I can't provide a "Fixes" tag.
>>>
>>> Hardly forever. It fixes da78dbff2e05630921c551dbbc70a4b7981a8fff.
>>
>> I don't see exactly how that can be true.
>>
>> That commit didn't change the parts of the NAPI poll processing which
>> are relevant here, mainly the guarding of the RX and TX work using
>> the status bits which are cleared.
>>
>
> AFAICS Francois is right and patch da78dbff2e05 ("r8169: remove work
> from irq handler") introduced the guarding of RX and TX work.
> I just checked back to 3.16 as oldest LTS kernel version.
Aha, now I see it.
>> Maybe I'm missing something? If so, indeed it would be nice to add
>> a proper Fixes: tag here.
>>
> Shall I submit a v2 including the Fixes line?
Yes, please do!
Powered by blists - more mailing lists