Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH net v1 2/2] lan743x: boost performance: limit PCIe
 bandwidth requirement
To:     Sven Van Asbroeck <thesven73@gmail.com>,
        Andrew Lunn <andrew@lunn.ch>
Cc:     Jakub Kicinski <kuba@kernel.org>,
        Bryan Whitehead <bryan.whitehead@microchip.com>,
        Microchip Linux Driver Support <UNGLinuxDriver@microchip.com>,
        David S Miller <davem@davemloft.net>,
        netdev <netdev@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
References: <20201206034408.31492-1-TheSven73@gmail.com>
 <20201206034408.31492-2-TheSven73@gmail.com>
 <20201208114314.743ee6ec@kicinski-fedora-pc1c0hjn.DHCP.thefacebook.com>
 <CAGngYiVSHRGC+eOCeF3Kyj_wOVqxJHvoc9fXRk-w+sVRjeSpcw@mail.gmail.com>
 <20201208225125.GA2602479@lunn.ch>
 <CAGngYiVp2u-A07rrkbeJCbqPW9efjkJUNC+NBxrtCM2JtXGpVA@mail.gmail.com>
From:   Florian Fainelli <f.fainelli@gmail.com>
Autocrypt: addr=f.fainelli@gmail.com; prefer-encrypt=mutual; keydata=
 mQGiBEjPuBIRBACW9MxSJU9fvEOCTnRNqG/13rAGsj+vJqontvoDSNxRgmafP8d3nesnqPyR
 xGlkaOSDuu09rxuW+69Y2f1TzjFuGpBk4ysWOR85O2Nx8AJ6fYGCoeTbovrNlGT1M9obSFGQ
 X3IzRnWoqlfudjTO5TKoqkbOgpYqIo5n1QbEjCCwCwCg3DOH/4ug2AUUlcIT9/l3pGvoRJ0E
 AICDzi3l7pmC5IWn2n1mvP5247urtHFs/uusE827DDj3K8Upn2vYiOFMBhGsxAk6YKV6IP0d
 ZdWX6fqkJJlu9cSDvWtO1hXeHIfQIE/xcqvlRH783KrihLcsmnBqOiS6rJDO2x1eAgC8meAX
 SAgsrBhcgGl2Rl5gh/jkeA5ykwbxA/9u1eEuL70Qzt5APJmqVXR+kWvrqdBVPoUNy/tQ8mYc
 nzJJ63ng3tHhnwHXZOu8hL4nqwlYHRa9eeglXYhBqja4ZvIvCEqSmEukfivk+DlIgVoOAJbh
 qIWgvr3SIEuR6ayY3f5j0f2ejUMYlYYnKdiHXFlF9uXm1ELrb0YX4GMHz7QnRmxvcmlhbiBG
 YWluZWxsaSA8Zi5mYWluZWxsaUBnbWFpbC5jb20+iGYEExECACYCGyMGCwkIBwMCBBUCCAME
 FgIDAQIeAQIXgAUCVF/S8QUJHlwd3wAKCRBhV5kVtWN2DvCVAJ4u4/bPF4P3jxb4qEY8I2gS
 6hG0gACffNWlqJ2T4wSSn+3o7CCZNd7SLSC5BA0ESM+4EhAQAL/o09boR9D3Vk1Tt7+gpYr3
 WQ6hgYVON905q2ndEoA2J0dQxJNRw3snabHDDzQBAcqOvdi7YidfBVdKi0wxHhSuRBfuOppu
 pdXkb7zxuPQuSveCLqqZWRQ+Cc2QgF7SBqgznbe6Ngout5qXY5Dcagk9LqFNGhJQzUGHAsIs
 hap1f0B1PoUyUNeEInV98D8Xd/edM3mhO9nRpUXRK9Bvt4iEZUXGuVtZLT52nK6Wv2EZ1TiT
 OiqZlf1P+vxYLBx9eKmabPdm3yjalhY8yr1S1vL0gSA/C6W1o/TowdieF1rWN/MYHlkpyj9c
 Rpc281gAO0AP3V1G00YzBEdYyi0gaJbCEQnq8Vz1vDXFxHzyhgGz7umBsVKmYwZgA8DrrB0M
 oaP35wuGR3RJcaG30AnJpEDkBYHznI2apxdcuTPOHZyEilIRrBGzDwGtAhldzlBoBwE3Z3MY
 31TOpACu1ZpNOMysZ6xiE35pWkwc0KYm4hJA5GFfmWSN6DniimW3pmdDIiw4Ifcx8b3mFrRO
 BbDIW13E51j9RjbO/nAaK9ndZ5LRO1B/8Fwat7bLzmsCiEXOJY7NNpIEpkoNoEUfCcZwmLrU
 +eOTPzaF6drw6ayewEi5yzPg3TAT6FV3oBsNg3xlwU0gPK3v6gYPX5w9+ovPZ1/qqNfOrbsE
 FRuiSVsZQ5s3AAMFD/9XjlnnVDh9GX/r/6hjmr4U9tEsM+VQXaVXqZuHKaSmojOLUCP/YVQo
 7IiYaNssCS4FCPe4yrL4FJJfJAsbeyDykMN7wAnBcOkbZ9BPJPNCbqU6dowLOiy8AuTYQ48m
 vIyQ4Ijnb6GTrtxIUDQeOBNuQC/gyyx3nbL/lVlHbxr4tb6YkhkO6shjXhQh7nQb33FjGO4P
 WU11Nr9i/qoV8QCo12MQEo244RRA6VMud06y/E449rWZFSTwGqb0FS0seTcYNvxt8PB2izX+
 HZA8SL54j479ubxhfuoTu5nXdtFYFj5Lj5x34LKPx7MpgAmj0H7SDhpFWF2FzcC1bjiW9mjW
 HaKaX23Awt97AqQZXegbfkJwX2Y53ufq8Np3e1542lh3/mpiGSilCsaTahEGrHK+lIusl6mz
 Joil+u3k01ofvJMK0ZdzGUZ/aPMZ16LofjFA+MNxWrZFrkYmiGdv+LG45zSlZyIvzSiG2lKy
 kuVag+IijCIom78P9jRtB1q1Q5lwZp2TLAJlz92DmFwBg1hyFzwDADjZ2nrDxKUiybXIgZp9
 aU2d++ptEGCVJOfEW4qpWCCLPbOT7XBr+g/4H3qWbs3j/cDDq7LuVYIe+wchy/iXEJaQVeTC
 y5arMQorqTFWlEOgRA8OP47L9knl9i4xuR0euV6DChDrguup2aJVU4hPBBgRAgAPAhsMBQJU
 X9LxBQkeXB3fAAoJEGFXmRW1Y3YOj4UAn3nrFLPZekMeqX5aD/aq/dsbXSfyAKC45Go0YyxV
 HGuUuzv+GKZ6nsysJ7kCDQRXG8fwARAA6q/pqBi5PjHcOAUgk2/2LR5LjjesK50bCaD4JuNc
 YDhFR7Vs108diBtsho3w8WRd9viOqDrhLJTroVckkk74OY8r+3t1E0Dd4wHWHQZsAeUvOwDM
 PQMqTUBFuMi6ydzTZpFA2wBR9x6ofl8Ax+zaGBcFrRlQnhsuXLnM1uuvS39+pmzIjasZBP2H
 UPk5ifigXcpelKmj6iskP3c8QN6x6GjUSmYx+xUfs/GNVSU1XOZn61wgPDbgINJd/THGdqiO
 iJxCLuTMqlSsmh1+E1dSdfYkCb93R/0ZHvMKWlAx7MnaFgBfsG8FqNtZu3PCLfizyVYYjXbV
 WO1A23riZKqwrSJAATo5iTS65BuYxrFsFNPrf7TitM8E76BEBZk0OZBvZxMuOs6Z1qI8YKVK
 UrHVGFq3NbuPWCdRul9SX3VfOunr9Gv0GABnJ0ET+K7nspax0xqq7zgnM71QEaiaH17IFYGS
 sG34V7Wo3vyQzsk7qLf9Ajno0DhJ+VX43g8+AjxOMNVrGCt9RNXSBVpyv2AMTlWCdJ5KI6V4
 KEzWM4HJm7QlNKE6RPoBxJVbSQLPd9St3h7mxLcne4l7NK9eNgNnneT7QZL8fL//s9K8Ns1W
 t60uQNYvbhKDG7+/yLcmJgjF74XkGvxCmTA1rW2bsUriM533nG9gAOUFQjURkwI8jvMAEQEA
 AYkCaAQYEQIACQUCVxvH8AIbAgIpCRBhV5kVtWN2DsFdIAQZAQIABgUCVxvH8AAKCRCH0Jac
 RAcHBIkHD/9nmfog7X2ZXMzL9ktT++7x+W/QBrSTCTmq8PK+69+INN1ZDOrY8uz6htfTLV9+
 e2W6G8/7zIvODuHk7r+yQ585XbplgP0V5Xc8iBHdBgXbqnY5zBrcH+Q/oQ2STalEvaGHqNoD
 UGyLQ/fiKoLZTPMur57Fy1c9rTuKiSdMgnT0FPfWVDfpR2Ds0gpqWePlRuRGOoCln5GnREA/
 2MW2rWf+CO9kbIR+66j8b4RUJqIK3dWn9xbENh/aqxfonGTCZQ2zC4sLd25DQA4w1itPo+f5
 V/SQxuhnlQkTOCdJ7b/mby/pNRz1lsLkjnXueLILj7gNjwTabZXYtL16z24qkDTI1x3g98R/
 xunb3/fQwR8FY5/zRvXJq5us/nLvIvOmVwZFkwXc+AF+LSIajqQz9XbXeIP/BDjlBNXRZNdo
 dVuSU51ENcMcilPr2EUnqEAqeczsCGpnvRCLfVQeSZr2L9N4svNhhfPOEscYhhpHTh0VPyxI
 pPBNKq+byuYPMyk3nj814NKhImK0O4gTyCK9b+gZAVvQcYAXvSouCnTZeJRrNHJFTgTgu6E0
 caxTGgc5zzQHeX67eMzrGomG3ZnIxmd1sAbgvJUDaD2GrYlulfwGWwWyTNbWRvMighVdPkSF
 6XFgQaosWxkV0OELLy2N485YrTr2Uq64VKyxpncLh50e2RnyAJ9Za0Dx0yyp44iD1OvHtkEI
 M5kY0ACeNhCZJvZ5g4C2Lc9fcTHu8jxmEkI=
Message-ID: <3aed88da-8e82-3bd0-6822-d30f1bd5ec9e@gmail.com>
Date:   Tue, 8 Dec 2020 15:36:20 -0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <CAGngYiVp2u-A07rrkbeJCbqPW9efjkJUNC+NBxrtCM2JtXGpVA@mail.gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 12/8/20 3:02 PM, Sven Van Asbroeck wrote:
> Hi Andrew,
> 
> On Tue, Dec 8, 2020 at 5:51 PM Andrew Lunn <andrew@lunn.ch> wrote:
>>
>>>
>>> So I assumed that it's a PCIe dma bandwidth issue, but I could be wrong -
>>> I didn't do any PCIe bandwidth measurements.
>>
>> Sometimes it is actually cache operations which take all the
>> time. This needs to invalidate the cache, so that when the memory is
>> then accessed, it get fetched from RAM. On SMP machines, cache
>> invalidation can be expensive, due to all the cross CPU operations.
>> I've actually got better performance by building a UP kernel on some
>> low core count ARM CPUs.
>>
>> There are some tricks which can be played. Do you actually need all
>> 9K? Does the descriptor tell you actually how much is used? You can
>> get a nice speed up if you just unmap 64 bytes for a TCP ACK, rather
>> than the full 9K.
>>
> 
> Thank you for the suggestion! The original driver developer chose 9K because
> presumably that's the largest frame size supported by the chip.
> 
> Yes, I believe the chip will tell us via the descriptor how much it has
> written, I would have to double-check. I was already looking for a
> "trick" to transfer only the required number of bytes, but I was led to
> believe that dma_map_single() and dma_unmap_single() always needed to match.
> 
> So:
> dma_map_single(9K) followed by dma_unmap_single(9K) is correct, and
> dma_map_single(9K) followed by dma_unmap_single(1500 bytes) means trouble.
> 
> How can we get around that?

dma_sync_single_for_{cpu,device} is what you would need in order to make
a partial cache line invalidation. You would still need to unmap the
same address+length pair that was used for the initial mapping otherwise
the DMA-API debugging will rightfully complain.
-- 
Florian