Linux-Amlogic Archive on lore.kernel.org
 help / Atom feed
* Re: stmmac / meson8b-dwmac
       [not found] ` <CAFBinCDebPOsmrhSXecx48nGWHh7g_OGPbr1Y0M+n_v9Ht91ew@mail.gmail.com>
@ 2019-01-17 21:23   ` Simon Huelck
  2019-02-04 14:34     ` Martin Blumenstingl
  0 siblings, 1 reply; 9+ messages in thread
From: Simon Huelck @ 2019-01-17 21:23 UTC (permalink / raw)
  To: martin.blumenstingl, linux-amlogic, netdev, Gpeppe.cavallaro,
	alexandre.torgue

Am 17.01.2019 um 21:57 schrieb Martin Blumenstingl:
> Hi Simon,
>
> On Thu, Jan 17, 2019 at 7:53 PM Simon Huelck <simonmail@gmx.de> wrote:
>> Hi Martin,
>>
>>
>> deutsch ?
> theoretisch: ja, das endet bei mir aber in einem komischen Mix aus
> deutsch und englisch, von daher...
>
>> I got problems with my ODROID c2 running on 4.19.16 ( and some releases
>> earlier ). the stmmac / dwmac driver doesnt provide the 800M/900M
>> performance that i was used to earlier.
>>
>>
>> Now im stuck near 550M/600M in the same environment. but what really
>> confuses me that duplex does hurt even more.
> interesting that you see this on the Odroid-C2 as well.
> previously I have only observed it on an Odroid-C1
>
>> PC --- VLAN3 --> switch --VLAN3--> ODROID
>>
>> NAS <-- VLAN1 -- switch <-- VLAN1-- ODROID
>>
>>
>> this means when im doing a iperf from PC to NAS, that my ODROID has load
>> on RX/TX same time (duplex). this shouldnt be an issue , all is 1GBits
>> FD. And in the past that wasnt an issue.
>>
>>
>> Now what happens:
>>
>> - benchmark between PC - ODROID is roughly 550M
>>
>> - benchmark between NAS - ODROID is roughly 550M
>>
>> - benchmark between PC - NAS is only around 300M
>>
>>
>> and like i said i was easliy able to hit 800 or even 900M to my NAS
>> earlier. I applied some .dtb fixes for interrupt levels for the
>> meson-gx.dtsi and meson-gxbb-odroid-c2.dtb, which will be mainlined ,
>> but the effect stayed identical.
> good that you have the interrupt patches already applied
> I believe it don't fix any performance issues - it's a fix for the
> Ethernet controller seemingly getting "stuck" (not processing data
> anymore). however, that already rules out one potential issue
>
>> are you aware of this problem ? Earlier kernel versions were all
>> perfectly fine and i stepped ( self compiled) kernel through all major
>> releases since odroid c2 was mainlined.
> I'm not aware of this - so it would be great if you could re-send your mail to:
> - the mainline Amlogic mailing list at: linux-amlogic@lists.infradead.org
> - the netdev mailing list (where all the networking people discuss
> their issues): netdev@vger.kernel.org
> - the stmmac maintainers: Gpeppe.cavallaro@st.com,
> alexandre.torgue@st.com, joabreu@synopsys.com


ok , will do so

>
> it's great that you stepped through various releases in the past.
> you can even help us to get closer to the root cause of the problem
> using git bisect. in case you haven't used git bisect yet::
> - git bisect start
> - git bisect good v4.18 (assuming v4.18 was good)
> - git bisect bad v4.19 (assuming v4.19 is bad)
> - the repeat the following:
> -- (git will checkout a different revision and ask you to test it)
> -- compile, boot the kernel and test whether your problem still exists
> -- enter either "git bisect good", "git bisect bad" or "git bisect
> skip" (for example if the revision doesn't compile, your board doesn't
> start with it, ...)
>
> git bisect will output one commit which is likely to be the cause (or
> at least a puzzle piece which points to the root cause)


the problem is that i dont have these kernel sources anymore :-(. but i
can provide some testing and numbers. maybe i dig if i got these kernel
configs somewhere around but i did not change much during migrating



im using a zyxel gs1900-8 switch and a qnap ts231p , and as i said i
didnt change my setup. i was able to hit 100MByte/s from my NAS , so
close to the benchmarks of 900MBit/s

>
>
> Regards
> Martin



_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-01-17 21:23   ` stmmac / meson8b-dwmac Simon Huelck
@ 2019-02-04 14:34     ` Martin Blumenstingl
  2019-02-06 10:36       ` Emiliano Ingrassia
  0 siblings, 1 reply; 9+ messages in thread
From: Martin Blumenstingl @ 2019-02-04 14:34 UTC (permalink / raw)
  To: ingrassia, Gpeppe.cavallaro, alexandre.torgue
  Cc: linux-amlogic, Simon Huelck, netdev

On Thu, Jan 17, 2019 at 10:23 PM Simon Huelck <simonmail@gmx.de> wrote:
[...]
> >> I got problems with my ODROID c2 running on 4.19.16 ( and some releases
> >> earlier ). the stmmac / dwmac driver doesnt provide the 800M/900M
> >> performance that i was used to earlier.
> >>
> >>
> >> Now im stuck near 550M/600M in the same environment. but what really
> >> confuses me that duplex does hurt even more.
> > interesting that you see this on the Odroid-C2 as well.
> > previously I have only observed it on an Odroid-C1
> >
> >> PC --- VLAN3 --> switch --VLAN3--> ODROID
> >>
> >> NAS <-- VLAN1 -- switch <-- VLAN1-- ODROID
> >>
> >>
> >> this means when im doing a iperf from PC to NAS, that my ODROID has load
> >> on RX/TX same time (duplex). this shouldnt be an issue , all is 1GBits
> >> FD. And in the past that wasnt an issue.
+Cc Emiliano who has seen a similar duplex issue on his Odroid-C1: [0]
(please note that all kernels prior to v5.1 with the pending patches
from [1] applied are only receiving data on RXD0 and RXD1 but not on
RXD2 and RXD3)

Emiliano, can you confirm the duplex issue observed by Simon is
similar to the one you see on your Odroid-C1?

> >>
> >>
> >> Now what happens:
> >>
> >> - benchmark between PC - ODROID is roughly 550M
> >>
> >> - benchmark between NAS - ODROID is roughly 550M
> >>
> >> - benchmark between PC - NAS is only around 300M
> >>
> >>
> >> and like i said i was easliy able to hit 800 or even 900M to my NAS
> >> earlier. I applied some .dtb fixes for interrupt levels for the
> >> meson-gx.dtsi and meson-gxbb-odroid-c2.dtb, which will be mainlined ,
> >> but the effect stayed identical.
> > good that you have the interrupt patches already applied
> > I believe it don't fix any performance issues - it's a fix for the
> > Ethernet controller seemingly getting "stuck" (not processing data
> > anymore). however, that already rules out one potential issue
> >
> >> are you aware of this problem ? Earlier kernel versions were all
> >> perfectly fine and i stepped ( self compiled) kernel through all major
> >> releases since odroid c2 was mainlined.
Guiseppe, Alexandre: what kind of data do you need from us if we see
the speeds drop (in both directions) when we send and receive at the
same time?

[...]
> the problem is that i dont have these kernel sources anymore :-(. but i
> can provide some testing and numbers. maybe i dig if i got these kernel
> configs somewhere around but i did not change much during migrating
do you remember the kernel version where it worked fine?

> im using a zyxel gs1900-8 switch and a qnap ts231p , and as i said i
> didnt change my setup. i was able to hit 100MByte/s from my NAS , so
> close to the benchmarks of 900MBit/s
I typically only do small transfers or I have traffic only in one direction.
thus it's likely that I missed this in my own tests


Regards
Martin


[0] http://lists.infradead.org/pipermail/linux-amlogic/2018-December/009679.html
[1] https://patchwork.kernel.org/cover/10744905/

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-02-04 14:34     ` Martin Blumenstingl
@ 2019-02-06 10:36       ` Emiliano Ingrassia
  2019-02-06 18:04         ` Simon Huelck
                           ` (2 more replies)
  0 siblings, 3 replies; 9+ messages in thread
From: Emiliano Ingrassia @ 2019-02-06 10:36 UTC (permalink / raw)
  To: Martin Blumenstingl, Simon Huelck
  Cc: linux-amlogic, netdev, alexandre.torgue, Gpeppe.cavallaro

Hi Martin, Hi Simon,

On Mon, Feb 04, 2019 at 03:34:41PM +0100, Martin Blumenstingl wrote:
> On Thu, Jan 17, 2019 at 10:23 PM Simon Huelck <simonmail@gmx.de> wrote:
> [...]
> > >> I got problems with my ODROID c2 running on 4.19.16 ( and some releases
> > >> earlier ). the stmmac / dwmac driver doesnt provide the 800M/900M
> > >> performance that i was used to earlier.
> > >>

Simon, did you ever reach 1 Gbps full duplex speed?
If yes, what was the kernel version did you use?

> > >>
> > >> Now im stuck near 550M/600M in the same environment. but what really
> > >> confuses me that duplex does hurt even more.
> > > interesting that you see this on the Odroid-C2 as well.
> > > previously I have only observed it on an Odroid-C1
> > >
> > >> PC --- VLAN3 --> switch --VLAN3--> ODROID
> > >>
> > >> NAS <-- VLAN1 -- switch <-- VLAN1-- ODROID
> > >>
> > >>
> > >> this means when im doing a iperf from PC to NAS, that my ODROID has load
> > >> on RX/TX same time (duplex). this shouldnt be an issue , all is 1GBits
> > >> FD. And in the past that wasnt an issue.
> +Cc Emiliano who has seen a similar duplex issue on his Odroid-C1: [0]
> (please note that all kernels prior to v5.1 with the pending patches
> from [1] applied are only receiving data on RXD0 and RXD1 but not on
> RXD2 and RXD3)
>
> Emiliano, can you confirm the duplex issue observed by Simon is
> similar to the one you see on your Odroid-C1?
>

It could be but, if I understand correctly, Simon is limited in
speed also in half duplex transmission (~550/600 Mbps), while we can
reach at least 900 Mbps.

> > >>
> > >>
> > >> Now what happens:
> > >>
> > >> - benchmark between PC - ODROID is roughly 550M
> > >>
> > >> - benchmark between NAS - ODROID is roughly 550M
> > >>
> > >> - benchmark between PC - NAS is only around 300M
> > >>
> > >>
> > >> and like i said i was easliy able to hit 800 or even 900M to my NAS
> > >> earlier. I applied some .dtb fixes for interrupt levels for the
> > >> meson-gx.dtsi and meson-gxbb-odroid-c2.dtb, which will be mainlined ,
> > >> but the effect stayed identical.
> > > good that you have the interrupt patches already applied
> > > I believe it don't fix any performance issues - it's a fix for the
> > > Ethernet controller seemingly getting "stuck" (not processing data
> > > anymore). however, that already rules out one potential issue
> > >
> > >> are you aware of this problem ? Earlier kernel versions were all
> > >> perfectly fine and i stepped ( self compiled) kernel through all major
> > >> releases since odroid c2 was mainlined.
> Guiseppe, Alexandre: what kind of data do you need from us if we see
> the speeds drop (in both directions) when we send and receive at the
> same time?
>
> [...]
> > the problem is that i dont have these kernel sources anymore :-(. but i
> > can provide some testing and numbers. maybe i dig if i got these kernel
> > configs somewhere around but i did not change much during migrating
> do you remember the kernel version where it worked fine?
>
> > im using a zyxel gs1900-8 switch and a qnap ts231p , and as i said i
> > didnt change my setup. i was able to hit 100MByte/s from my NAS , so
> > close to the benchmarks of 900MBit/s
> I typically only do small transfers or I have traffic only in one direction.
> thus it's likely that I missed this in my own tests
>
>
> Regards
> Martin
>
>
> [0] http://lists.infradead.org/pipermail/linux-amlogic/2018-December/009679.html
> [1] https://patchwork.kernel.org/cover/10744905/

Regards,

Emiliano

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-02-06 10:36       ` Emiliano Ingrassia
@ 2019-02-06 18:04         ` Simon Huelck
  2019-02-06 21:21         ` Simon Huelck
  2019-02-07 19:30         ` Simon Huelck
  2 siblings, 0 replies; 9+ messages in thread
From: Simon Huelck @ 2019-02-06 18:04 UTC (permalink / raw)
  To: Emiliano Ingrassia, Martin Blumenstingl
  Cc: linux-amlogic, netdev, alexandre.torgue, Gpeppe.cavallaro

Hi,


yes i did reach 900MBits . maybe it was kernel 4.14 or 4.17. The 3.x
series was also ok.

I will pull latest 4.17 and test. If this doesnt work i go for 4.14

regards,
Simon

Am 06.02.2019 um 11:36 schrieb Emiliano Ingrassia:
> Hi Martin, Hi Simon,
>
> On Mon, Feb 04, 2019 at 03:34:41PM +0100, Martin Blumenstingl wrote:
>> On Thu, Jan 17, 2019 at 10:23 PM Simon Huelck <simonmail@gmx.de> wrote:
>> [...]
>>>>> I got problems with my ODROID c2 running on 4.19.16 ( and some releases
>>>>> earlier ). the stmmac / dwmac driver doesnt provide the 800M/900M
>>>>> performance that i was used to earlier.
>>>>>
> Simon, did you ever reach 1 Gbps full duplex speed?
> If yes, what was the kernel version did you use?
>
>>>>> Now im stuck near 550M/600M in the same environment. but what really
>>>>> confuses me that duplex does hurt even more.
>>>> interesting that you see this on the Odroid-C2 as well.
>>>> previously I have only observed it on an Odroid-C1
>>>>
>>>>> PC --- VLAN3 --> switch --VLAN3--> ODROID
>>>>>
>>>>> NAS <-- VLAN1 -- switch <-- VLAN1-- ODROID
>>>>>
>>>>>
>>>>> this means when im doing a iperf from PC to NAS, that my ODROID has load
>>>>> on RX/TX same time (duplex). this shouldnt be an issue , all is 1GBits
>>>>> FD. And in the past that wasnt an issue.
>> +Cc Emiliano who has seen a similar duplex issue on his Odroid-C1: [0]
>> (please note that all kernels prior to v5.1 with the pending patches
>> from [1] applied are only receiving data on RXD0 and RXD1 but not on
>> RXD2 and RXD3)
>>
>> Emiliano, can you confirm the duplex issue observed by Simon is
>> similar to the one you see on your Odroid-C1?
>>
> It could be but, if I understand correctly, Simon is limited in
> speed also in half duplex transmission (~550/600 Mbps), while we can
> reach at least 900 Mbps.
>
>>>>>
>>>>> Now what happens:
>>>>>
>>>>> - benchmark between PC - ODROID is roughly 550M
>>>>>
>>>>> - benchmark between NAS - ODROID is roughly 550M
>>>>>
>>>>> - benchmark between PC - NAS is only around 300M
>>>>>
>>>>>
>>>>> and like i said i was easliy able to hit 800 or even 900M to my NAS
>>>>> earlier. I applied some .dtb fixes for interrupt levels for the
>>>>> meson-gx.dtsi and meson-gxbb-odroid-c2.dtb, which will be mainlined ,
>>>>> but the effect stayed identical.
>>>> good that you have the interrupt patches already applied
>>>> I believe it don't fix any performance issues - it's a fix for the
>>>> Ethernet controller seemingly getting "stuck" (not processing data
>>>> anymore). however, that already rules out one potential issue
>>>>
>>>>> are you aware of this problem ? Earlier kernel versions were all
>>>>> perfectly fine and i stepped ( self compiled) kernel through all major
>>>>> releases since odroid c2 was mainlined.
>> Guiseppe, Alexandre: what kind of data do you need from us if we see
>> the speeds drop (in both directions) when we send and receive at the
>> same time?
>>
>> [...]
>>> the problem is that i dont have these kernel sources anymore :-(. but i
>>> can provide some testing and numbers. maybe i dig if i got these kernel
>>> configs somewhere around but i did not change much during migrating
>> do you remember the kernel version where it worked fine?
>>
>>> im using a zyxel gs1900-8 switch and a qnap ts231p , and as i said i
>>> didnt change my setup. i was able to hit 100MByte/s from my NAS , so
>>> close to the benchmarks of 900MBit/s
>> I typically only do small transfers or I have traffic only in one direction.
>> thus it's likely that I missed this in my own tests
>>
>>
>> Regards
>> Martin
>>
>>
>> [0] http://lists.infradead.org/pipermail/linux-amlogic/2018-December/009679.html
>> [1] https://patchwork.kernel.org/cover/10744905/
> Regards,
>
> Emiliano



_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-02-06 10:36       ` Emiliano Ingrassia
  2019-02-06 18:04         ` Simon Huelck
@ 2019-02-06 21:21         ` Simon Huelck
  2019-02-07 19:30         ` Simon Huelck
  2 siblings, 0 replies; 9+ messages in thread
From: Simon Huelck @ 2019-02-06 21:21 UTC (permalink / raw)
  To: Emiliano Ingrassia, Martin Blumenstingl
  Cc: linux-amlogic, netdev, alexandre.torgue, Gpeppe.cavallaro

Hi,

4.17.9 failed going back further

regards,
Simon

Am 06.02.2019 um 11:36 schrieb Emiliano Ingrassia:
> Hi Martin, Hi Simon,
>
> On Mon, Feb 04, 2019 at 03:34:41PM +0100, Martin Blumenstingl wrote:
>> On Thu, Jan 17, 2019 at 10:23 PM Simon Huelck <simonmail@gmx.de> wrote:
>> [...]
>>>>> I got problems with my ODROID c2 running on 4.19.16 ( and some releases
>>>>> earlier ). the stmmac / dwmac driver doesnt provide the 800M/900M
>>>>> performance that i was used to earlier.
>>>>>
> Simon, did you ever reach 1 Gbps full duplex speed?
> If yes, what was the kernel version did you use?
>
>>>>> Now im stuck near 550M/600M in the same environment. but what really
>>>>> confuses me that duplex does hurt even more.
>>>> interesting that you see this on the Odroid-C2 as well.
>>>> previously I have only observed it on an Odroid-C1
>>>>
>>>>> PC --- VLAN3 --> switch --VLAN3--> ODROID
>>>>>
>>>>> NAS <-- VLAN1 -- switch <-- VLAN1-- ODROID
>>>>>
>>>>>
>>>>> this means when im doing a iperf from PC to NAS, that my ODROID has load
>>>>> on RX/TX same time (duplex). this shouldnt be an issue , all is 1GBits
>>>>> FD. And in the past that wasnt an issue.
>> +Cc Emiliano who has seen a similar duplex issue on his Odroid-C1: [0]
>> (please note that all kernels prior to v5.1 with the pending patches
>> from [1] applied are only receiving data on RXD0 and RXD1 but not on
>> RXD2 and RXD3)
>>
>> Emiliano, can you confirm the duplex issue observed by Simon is
>> similar to the one you see on your Odroid-C1?
>>
> It could be but, if I understand correctly, Simon is limited in
> speed also in half duplex transmission (~550/600 Mbps), while we can
> reach at least 900 Mbps.
>
>>>>>
>>>>> Now what happens:
>>>>>
>>>>> - benchmark between PC - ODROID is roughly 550M
>>>>>
>>>>> - benchmark between NAS - ODROID is roughly 550M
>>>>>
>>>>> - benchmark between PC - NAS is only around 300M
>>>>>
>>>>>
>>>>> and like i said i was easliy able to hit 800 or even 900M to my NAS
>>>>> earlier. I applied some .dtb fixes for interrupt levels for the
>>>>> meson-gx.dtsi and meson-gxbb-odroid-c2.dtb, which will be mainlined ,
>>>>> but the effect stayed identical.
>>>> good that you have the interrupt patches already applied
>>>> I believe it don't fix any performance issues - it's a fix for the
>>>> Ethernet controller seemingly getting "stuck" (not processing data
>>>> anymore). however, that already rules out one potential issue
>>>>
>>>>> are you aware of this problem ? Earlier kernel versions were all
>>>>> perfectly fine and i stepped ( self compiled) kernel through all major
>>>>> releases since odroid c2 was mainlined.
>> Guiseppe, Alexandre: what kind of data do you need from us if we see
>> the speeds drop (in both directions) when we send and receive at the
>> same time?
>>
>> [...]
>>> the problem is that i dont have these kernel sources anymore :-(. but i
>>> can provide some testing and numbers. maybe i dig if i got these kernel
>>> configs somewhere around but i did not change much during migrating
>> do you remember the kernel version where it worked fine?
>>
>>> im using a zyxel gs1900-8 switch and a qnap ts231p , and as i said i
>>> didnt change my setup. i was able to hit 100MByte/s from my NAS , so
>>> close to the benchmarks of 900MBit/s
>> I typically only do small transfers or I have traffic only in one direction.
>> thus it's likely that I missed this in my own tests
>>
>>
>> Regards
>> Martin
>>
>>
>> [0] http://lists.infradead.org/pipermail/linux-amlogic/2018-December/009679.html
>> [1] https://patchwork.kernel.org/cover/10744905/
> Regards,
>
> Emiliano



_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-02-06 10:36       ` Emiliano Ingrassia
  2019-02-06 18:04         ` Simon Huelck
  2019-02-06 21:21         ` Simon Huelck
@ 2019-02-07 19:30         ` Simon Huelck
  2019-02-09  1:09           ` Martin Blumenstingl
  2 siblings, 1 reply; 9+ messages in thread
From: Simon Huelck @ 2019-02-07 19:30 UTC (permalink / raw)
  To: Emiliano Ingrassia, Martin Blumenstingl
  Cc: linux-amlogic, netdev, alexandre.torgue, Gpeppe.cavallaro

Hi Guys,


i can confirm better performance with 4.14.29

- ~900 MBits with iperf2 in one way
-~ 500 - 600MBits with iperf2 in duplex in both directions


This wasnt the case with 4.17.9, not with 4.18, 4.19 or the 5.0 series.....


How can i help further ?

regards,
Simon


Am 06.02.2019 um 11:36 schrieb Emiliano Ingrassia:
> Hi Martin, Hi Simon,
>
> On Mon, Feb 04, 2019 at 03:34:41PM +0100, Martin Blumenstingl wrote:
>> On Thu, Jan 17, 2019 at 10:23 PM Simon Huelck <simonmail@gmx.de> wrote:
>> [...]
>>>>> I got problems with my ODROID c2 running on 4.19.16 ( and some releases
>>>>> earlier ). the stmmac / dwmac driver doesnt provide the 800M/900M
>>>>> performance that i was used to earlier.
>>>>>
> Simon, did you ever reach 1 Gbps full duplex speed?
> If yes, what was the kernel version did you use?
>
>>>>> Now im stuck near 550M/600M in the same environment. but what really
>>>>> confuses me that duplex does hurt even more.
>>>> interesting that you see this on the Odroid-C2 as well.
>>>> previously I have only observed it on an Odroid-C1
>>>>
>>>>> PC --- VLAN3 --> switch --VLAN3--> ODROID
>>>>>
>>>>> NAS <-- VLAN1 -- switch <-- VLAN1-- ODROID
>>>>>
>>>>>
>>>>> this means when im doing a iperf from PC to NAS, that my ODROID has load
>>>>> on RX/TX same time (duplex). this shouldnt be an issue , all is 1GBits
>>>>> FD. And in the past that wasnt an issue.
>> +Cc Emiliano who has seen a similar duplex issue on his Odroid-C1: [0]
>> (please note that all kernels prior to v5.1 with the pending patches
>> from [1] applied are only receiving data on RXD0 and RXD1 but not on
>> RXD2 and RXD3)
>>
>> Emiliano, can you confirm the duplex issue observed by Simon is
>> similar to the one you see on your Odroid-C1?
>>
> It could be but, if I understand correctly, Simon is limited in
> speed also in half duplex transmission (~550/600 Mbps), while we can
> reach at least 900 Mbps.
>
>>>>>
>>>>> Now what happens:
>>>>>
>>>>> - benchmark between PC - ODROID is roughly 550M
>>>>>
>>>>> - benchmark between NAS - ODROID is roughly 550M
>>>>>
>>>>> - benchmark between PC - NAS is only around 300M
>>>>>
>>>>>
>>>>> and like i said i was easliy able to hit 800 or even 900M to my NAS
>>>>> earlier. I applied some .dtb fixes for interrupt levels for the
>>>>> meson-gx.dtsi and meson-gxbb-odroid-c2.dtb, which will be mainlined ,
>>>>> but the effect stayed identical.
>>>> good that you have the interrupt patches already applied
>>>> I believe it don't fix any performance issues - it's a fix for the
>>>> Ethernet controller seemingly getting "stuck" (not processing data
>>>> anymore). however, that already rules out one potential issue
>>>>
>>>>> are you aware of this problem ? Earlier kernel versions were all
>>>>> perfectly fine and i stepped ( self compiled) kernel through all major
>>>>> releases since odroid c2 was mainlined.
>> Guiseppe, Alexandre: what kind of data do you need from us if we see
>> the speeds drop (in both directions) when we send and receive at the
>> same time?
>>
>> [...]
>>> the problem is that i dont have these kernel sources anymore :-(. but i
>>> can provide some testing and numbers. maybe i dig if i got these kernel
>>> configs somewhere around but i did not change much during migrating
>> do you remember the kernel version where it worked fine?
>>
>>> im using a zyxel gs1900-8 switch and a qnap ts231p , and as i said i
>>> didnt change my setup. i was able to hit 100MByte/s from my NAS , so
>>> close to the benchmarks of 900MBit/s
>> I typically only do small transfers or I have traffic only in one direction.
>> thus it's likely that I missed this in my own tests
>>
>>
>> Regards
>> Martin
>>
>>
>> [0] http://lists.infradead.org/pipermail/linux-amlogic/2018-December/009679.html
>> [1] https://patchwork.kernel.org/cover/10744905/
> Regards,
>
> Emiliano



_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-02-07 19:30         ` Simon Huelck
@ 2019-02-09  1:09           ` Martin Blumenstingl
  2019-02-11 13:44             ` Jose Abreu
  0 siblings, 1 reply; 9+ messages in thread
From: Martin Blumenstingl @ 2019-02-09  1:09 UTC (permalink / raw)
  To: Simon Huelck
  Cc: linux-amlogic, netdev, alexandre.torgue, Emiliano Ingrassia,
	Gpeppe.cavallaro

Hi Simon,

On Thu, Feb 7, 2019 at 8:30 PM Simon Huelck <simonmail@gmx.de> wrote:
>
> Hi Guys,
>
>
> i can confirm better performance with 4.14.29
>
> - ~900 MBits with iperf2 in one way
> -~ 500 - 600MBits with iperf2 in duplex in both directions
>
>
> This wasnt the case with 4.17.9, not with 4.18, 4.19 or the 5.0 series.....
I just did a small test myself on a Khadas VIM2:
# iperf3 -c 192.168.1.100
Connecting to host 192.168.1.100, port 5201
[  5] local 192.168.1.189 port 37192 connected to 192.168.1.100 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec   113 MBytes   946 Mbits/sec    0    354 KBytes
[  5]   1.00-2.00   sec   112 MBytes   940 Mbits/sec    0    354 KBytes
[  5]   2.00-3.00   sec   110 MBytes   920 Mbits/sec  241    228 KBytes
[  5]   3.00-4.00   sec   112 MBytes   940 Mbits/sec    0    314 KBytes
[  5]   4.00-5.00   sec   111 MBytes   933 Mbits/sec   89   83.4 KBytes
[  5]   5.00-6.00   sec   110 MBytes   926 Mbits/sec  115    335 KBytes
[  5]   6.00-7.00   sec   112 MBytes   941 Mbits/sec    0    358 KBytes
[  5]   7.00-8.00   sec   112 MBytes   941 Mbits/sec    0    362 KBytes
[  5]   8.00-9.00   sec   112 MBytes   941 Mbits/sec    0    369 KBytes
[  5]   9.00-10.00  sec   112 MBytes   942 Mbits/sec    0    372 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  1.09 GBytes   937 Mbits/sec  445             sender
[  5]   0.00-10.04  sec  1.09 GBytes   932 Mbits/sec                  receiver

iperf Done.

(it's interesting that the sending direction has 445 retries)

# iperf3 -c 192.168.1.100 -R
Connecting to host 192.168.1.100, port 5201
Reverse mode, remote host 192.168.1.100 is sending
[  5] local 192.168.1.189 port 37196 connected to 192.168.1.100 port 5201
[ ID] Interval           Transfer     Bitrate
[  5]   0.00-1.00   sec  90.9 MBytes   763 Mbits/sec
[  5]   1.00-2.00   sec  90.9 MBytes   762 Mbits/sec
[  5]   2.00-3.00   sec  90.7 MBytes   760 Mbits/sec
[  5]   3.00-4.00   sec  91.3 MBytes   766 Mbits/sec
[  5]   4.00-5.00   sec  91.1 MBytes   764 Mbits/sec
[  5]   5.00-6.00   sec  91.1 MBytes   765 Mbits/sec
[  5]   6.00-7.00   sec  90.8 MBytes   762 Mbits/sec
[  5]   7.00-8.00   sec  90.9 MBytes   762 Mbits/sec
[  5]   8.00-9.00   sec  91.0 MBytes   764 Mbits/sec
[  5]   9.00-10.00  sec  91.3 MBytes   766 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.04  sec   911 MBytes   762 Mbits/sec    0             sender
[  5]   0.00-10.00  sec   910 MBytes   763 Mbits/sec                  receiver

iperf Done.

(when receiving I see no retries)

for my test I used my Khadas VIM2 (as I don't have a GXBB board anymore).
test setup: PC -> built-in switch in some ath79 based OpenWrt device
-> VIM2. no VLANs are used
revision: latest mainline, which at the time of testing is:
46c291e277f937378 ("Merge tag 'armsoc-fixes-5.0' of
git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc")

> How can i help further ?
it's good to know that 4.14 has "good" performance in your scenario
can you please show the full iperf outputs for your tests (preferably
on both, 4.14 and 5.0-rcX)?

do you see any improvements on 5.0-rcX when not using VLANs (this is
just a random guess)?


Regards
Martin

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-02-09  1:09           ` Martin Blumenstingl
@ 2019-02-11 13:44             ` Jose Abreu
  2019-02-14  7:21               ` Simon Huelck
  0 siblings, 1 reply; 9+ messages in thread
From: Jose Abreu @ 2019-02-11 13:44 UTC (permalink / raw)
  To: Martin Blumenstingl, Simon Huelck
  Cc: linux-amlogic, netdev, alexandre.torgue, Emiliano Ingrassia,
	Gpeppe.cavallaro

Hello,

On 2/9/2019 1:09 AM, Martin Blumenstingl wrote:
> (it's interesting that the sending direction has 445 retries)

I saw this before and I think it was related with COE. Can you
please disable all offloading and try again?

Anyway, maybe Simon should bissect ? I think since 4.14 there are
not that many commits in stmmac / meson8b-dwmac that can cause
this behavior.

Thanks,
Jose Miguel Abreu

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: stmmac / meson8b-dwmac
  2019-02-11 13:44             ` Jose Abreu
@ 2019-02-14  7:21               ` Simon Huelck
  0 siblings, 0 replies; 9+ messages in thread
From: Simon Huelck @ 2019-02-14  7:21 UTC (permalink / raw)
  To: Jose Abreu, Martin Blumenstingl
  Cc: linux-amlogic, netdev, alexandre.torgue, Emiliano Ingrassia,
	Gpeppe.cavallaro

Hi,

i used iperf2 on my odroid c2 , since this can do duplex tests ! iperf2 -c 10.10.11.1
-i1 -d


Do you really need the exact numbers ?

NON-duplex:
4.14.29 reached 930 MBits without the -d ( duplex option ), 5.0rc5 only
reaches 600MBits there

DUPLEX:
With the -d option things get worse on 4.14.29 , 600MBits in both
directions. on 5.0rc5 this is the 500 to 600 in one direction, the other
only 150MBits.


iperf3 is not duplex capable, i didnt test the reverse option, but
duplex tests show that one direction drops in performance when the other
is busy.


Do you really need the exact output ? I need to reflash my box then
..... if needed i would do so.

regards,
Simon


Am 11.02.2019 um 14:44 schrieb Jose Abreu:
> Hello,
>
> On 2/9/2019 1:09 AM, Martin Blumenstingl wrote:
>> (it's interesting that the sending direction has 445 retries)
> I saw this before and I think it was related with COE. Can you
> please disable all offloading and try again?
>
> Anyway, maybe Simon should bissect ? I think since 4.14 there are
> not that many commits in stmmac / meson8b-dwmac that can cause
> this behavior.
>
> Thanks,
> Jose Miguel Abreu



_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, back to index

Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <a38e643c-ed9f-c306-cc95-84f70ebc1f10@gmx.de>
     [not found] ` <CAFBinCDebPOsmrhSXecx48nGWHh7g_OGPbr1Y0M+n_v9Ht91ew@mail.gmail.com>
2019-01-17 21:23   ` stmmac / meson8b-dwmac Simon Huelck
2019-02-04 14:34     ` Martin Blumenstingl
2019-02-06 10:36       ` Emiliano Ingrassia
2019-02-06 18:04         ` Simon Huelck
2019-02-06 21:21         ` Simon Huelck
2019-02-07 19:30         ` Simon Huelck
2019-02-09  1:09           ` Martin Blumenstingl
2019-02-11 13:44             ` Jose Abreu
2019-02-14  7:21               ` Simon Huelck

Linux-Amlogic Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-amlogic/0 linux-amlogic/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-amlogic linux-amlogic/ https://lore.kernel.org/linux-amlogic \
		linux-amlogic@lists.infradead.org linux-amlogic@archiver.kernel.org
	public-inbox-index linux-amlogic


Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.infradead.lists.linux-amlogic


AGPL code for this site: git clone https://public-inbox.org/ public-inbox