netdev - Re: [PATCH net-next v3 08/12] net: ethernet: oa_tc6: implement transmit path to transfer tx ethernet frames

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <53b090b1-d7bb-4a81-9f0b-9979db8dec59@microchip.com>
Date: Wed, 20 Mar 2024 10:43:18 +0000
From: <Parthiban.Veerasooran@...rochip.com>
To: <andrew@...n.ch>
CC: <davem@...emloft.net>, <edumazet@...gle.com>, <kuba@...nel.org>,
	<pabeni@...hat.com>, <horms@...nel.org>, <saeedm@...dia.com>,
	<anthony.l.nguyen@...el.com>, <netdev@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, <corbet@....net>,
	<linux-doc@...r.kernel.org>, <robh+dt@...nel.org>,
	<krzysztof.kozlowski+dt@...aro.org>, <conor+dt@...nel.org>,
	<devicetree@...r.kernel.org>, <Horatiu.Vultur@...rochip.com>,
	<ruanjinjie@...wei.com>, <Steen.Hegelund@...rochip.com>,
	<vladimir.oltean@....com>, <UNGLinuxDriver@...rochip.com>,
	<Thorsten.Kummermehr@...rochip.com>, <Pier.Beruto@...emi.com>,
	<Selvamani.Rajagopal@...emi.com>, <Nicolas.Ferre@...rochip.com>,
	<benjamin.bigler@...nformulastudent.ch>
Subject: Re: [PATCH net-next v3 08/12] net: ethernet: oa_tc6: implement
 transmit path to transfer tx ethernet frames

Hi Andrew,

On 19/03/24 6:49 pm, Andrew Lunn wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> On Tue, Mar 19, 2024 at 12:54:30PM +0000, Parthiban.Veerasooran@...rochip.com wrote:
>> Hi Andrew,
>>
>> On 07/03/24 10:38 pm, Andrew Lunn wrote:
>>> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
>>>
>>>> @@ -55,6 +77,14 @@
>>>>                                                 (OA_TC6_CTRL_MAX_REGISTERS *\
>>>>                                                 OA_TC6_CTRL_REG_VALUE_SIZE) +\
>>>>                                                 OA_TC6_CTRL_IGNORED_SIZE)
>>>> +#define OA_TC6_CHUNK_PAYLOAD_SIZE            64
>>>> +#define OA_TC6_DATA_HEADER_SIZE                      4
>>>> +#define OA_TC6_CHUNK_SIZE                    (OA_TC6_DATA_HEADER_SIZE +\
>>>> +                                             OA_TC6_CHUNK_PAYLOAD_SIZE)
>>>> +#define OA_TC6_TX_SKB_QUEUE_SIZE             100
>>>
>>> So you keep up to 100 packets in a queue. If use assume typical MTU
>>> size packets, that is 1,238,400 bits. At 10Mbps, that is 120ms of
>>> traffic. That is quite a lot of latency when a high priority packet is
>>> added to the tail of the queue and needs to wait for all the other
>>> packets to be sent first.
>>>
>>> Chunks are 64 bytes. So in practice, you only ever need two
>>> packets. You need to be able to fill a chunk with the final part of
>>> one packet, and the beginning of the next. So i would try using a much
>>> smaller queue size. That will allow Linux queue disciplines to give
>>> you the high priority packets first which you send with low latency.
>> Thanks for the detailed explanation. If I understand you correctly,
>>
>> 1. The tx skb queue size (OA_TC6_TX_SKB_QUEUE_SIZE) should be 2 to avoid
>> the latency when a high priority packet added.
>>
>> 2. Need to implement the handling part of the below case,
>> In case if one packet ends in a chunk and that chunk still having some
>> space left to accommodate some bytes from the next packet if available
>> from network layer.
> 
> This second part is clearly an optimisation. If you have lots of full
> MTU packets, 1514 bytes, they take around 24 chunks. Having the last
> chunk only 1/2 full does not waste too much bandwidth. But if you are
> carrying lots of small packets, say voice, 130 bytes, the wasted
> bandwidth starts to add up. But is there a use case for 10Mbps of
> small packets? I doubt it.
Yes, for sure there is a possibility to get into this scenario and the 
protocol also supports that. But as proposed by you below, let's 
implement it as part of optimization later.
> 
> So if you don't have the ability to combine two packets into one
> chunk, i would do that later. Lets get the basics merged first, it can
> be optimised later.
Yes, I agree with this proposal to get the basic version merged first.

Best regards,
Parthiban V
> 
>          Andrew