QCA6174 Performance

QCA6174 Performance

[Daniel Holz]

Hello,

the network performance of the Qualcom QCA6174 in my Lenovo Yoga 3 11 is 
only around 7-8 Mbyte/s in tests with iperf when using a wireless ac 
connection. Disabling power management resulted in 10 MByte/s. Under the 
same conditions my Android phone achieved 384 MByte/s.

Currently I'm using Ubuntu 16.04 on my Yoga 3 11 with the most recent 
hw2.1 firmware and kernel 4.4. I already tried kernel 4.8 which boosted 
the speed to 15 MByte/s, but this is still way too slow for such a 
connection.


Thanks

Daniel Holz


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: QCA6174 Performance

[Valo, Kalle]

Daniel Holz <daniel.holz91@gmx.de> writes:

> the network performance of the Qualcom QCA6174 in my Lenovo Yoga 3 11
> is only around 7-8 Mbyte/s in tests with iperf when using a wireless
> ac connection. Disabling power management resulted in 10 MByte/s.
> Under the same conditions my Android phone achieved 384 MByte/s.
>
> Currently I'm using Ubuntu 16.04 on my Yoga 3 11 with the most recent
> hw2.1 firmware and kernel 4.4. I already tried kernel 4.8 which
> boosted the speed to 15 MByte/s, but this is still way too slow for
> such a connection.

Might be a problem with the board file, are you using the latest one?
dmesg and 'lspci -vnn' outputs help to analyse the issue more.

-- 
Kalle Valo
_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: QCA6174 Performance

[Valo, Kalle]

Hi,

please don't email me privately, I'm adding the mailing list back so
that everyone can participate.

Daniel Holz <daniel.holz91@gmail.com> writes:

> Do you mean by board file board.bin? I downloaded all files related to the
> QCA6174 from your Github repository.

I meant board-2.bin actually:

$ md5sum QCA6174/hw2.1/board-2.bin 
11920f7c30a1b1dd009928baef1021dc  QCA6174/hw2.1/board-2.bin

This file should support your board with subsystem id 17aa:3044.

> Here is the output of lspci:
>
> 02:00.0 Network controller [0280]: Qualcomm Atheros QCA6174 802.11ac Wireless
> Network Adapter [168c:003e] (rev 20)
> Subsystem: Lenovo QCA6174 802.11ac Wireless Network Adapter [17aa:3044]
> Flags: bus master, fast devsel, latency 0, IRQ 49
> Memory at c1000000 (64-bit, non-prefetchable) [size=2M]
> Capabilities: [40] Power Management version 3
> Capabilities: [50] MSI: Enable+ Count=8/8 Maskable+ 64bit-
> Capabilities: [70] Express Endpoint, MSI 00
> Capabilities: [100] Advanced Error Reporting
> Capabilities: [148] Virtual Channel
> Capabilities: [168] Device Serial Number 00-00-00-00-00-00-00-00
> Capabilities: [178] Latency Tolerance Reporting
> Capabilities: [180] L1 PM Substates
> Kernel driver in use: ath10k_pci
> Kernel modules: ath10k_pci
>
> And of dmesg | grep -i ath10k:
>
> [    3.717494] ath10k_pci 0000:02:00.0: pci irq msi-x interrupts 8 irq_mode 0
> reset_mode 0
> [    3.953007] ath10k_pci 0000:02:00.0: Direct firmware load for ath10k/
> cal-pci-0000:02:00.0.bin failed with error -2
> [    5.232572] ath10k_pci 0000:02:00.0: qca6174 hw2.1 (0x05010000, 0x003405ff
> sub 17aa:3044) fw SW_RM.1.1.1-00157-QCARMSWPZ-1 fwapi 5 bdapi 2 htt-ver 3.1
> wmi-op 4 htt-op 3 cal otp max-sta 32 raw 0 hwcrypto 1 features
> ignore-otp,no-4addr-pad
> [    5.232578] ath10k_pci 0000:02:00.0: debug 0 debugfs 1 tracing 1 dfs 0
> testmode 0
> [    5.318496] ath10k_pci 0000:02:00.0 wlp2s0: renamed from wlan0
> [    6.879214] ath10k_pci 0000:02:00.0: no channel configured; ignoring frame
> (s)!

From the logs I can't immeaditely see anything wrong, but you version of
ath10k is so old that it doesn't print the crc32 for the board file so I
can't be 100% sure. So I recommend to check the md5sum above matches
with your file.

From a quick look I can't immediately come up what else would cause low
throughput, it would involve more detailed analysis with wireless
sniffer etc. It's a wild shot but you could always try latest ath10k
using kernel-backports:

https://wireless.wiki.kernel.org/en/users/drivers/ath10k/backports

Do note that backports project hasn't had stable updates for a long time
so use the daily releases instead:

https://www.kernel.org/pub/linux/kernel/projects/backports/2016/03/24/

-- 
Kalle Valo
_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: QCA6174 Performance

[Daniel Holz]

iw wlp2s0 info displays that my bandwidth is 80 MHz. Over the weekend I 
tried different configurations like using n instead of ac or 2,4 GHz 
and connection speed on a different WLAN, but the performance was 
always extremely slow.

Daniel Holz

Am Fr, 16. Sep, 2016 um 2:23 schrieb Michal Kazior 
<michal.kazior@tieto.com>:
> On 16 September 2016 at 13:50, Daniel Holz <daniel.holz91@gmx.de> 
> wrote:
>>  The md5sum matches.
>> 
>> 
>>  I installed Kernel 4.8 from Ubuntu's mainline ppa to get the recent 
>> version
>>  of ath10k which increased the performance, but it's still pretty 
>> low.
>> 
>>  Here is the output of dmesg with kernel 4.8:
> 
> You can check it with a recent enough iw:
>   iw wlp2s0 info
> 
> It should tell you channel bandwidth you are using.
> 
> Maybe ath10k or your AP has a bug that causes wrong capabilities to be
> negotiated during association.
> 
> You could try re-configuring your AP (to slower speeds) or force
> wpa_supplicant/ath10k to not use vht or ht modes at all and re-testing
> if it helps.
> 
> 
> Michał


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: QCA6174 Performance

[Michal Kazior]

On 16 September 2016 at 13:50, Daniel Holz <daniel.holz91@gmx.de> wrote:
> The md5sum matches.
>
>
> I installed Kernel 4.8 from Ubuntu's mainline ppa to get the recent version
> of ath10k which increased the performance, but it's still pretty low.
>
> Here is the output of dmesg with kernel 4.8:

You can check it with a recent enough iw:
  iw wlp2s0 info

It should tell you channel bandwidth you are using.

Maybe ath10k or your AP has a bug that causes wrong capabilities to be
negotiated during association.

You could try re-configuring your AP (to slower speeds) or force
wpa_supplicant/ath10k to not use vht or ht modes at all and re-testing
if it helps.


Michał

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: QCA6174 Performance

[Daniel Holz]

The md5sum matches.


I installed Kernel 4.8 from Ubuntu's mainline ppa to get the recent 
version of ath10k which increased the performance, but it's still pretty 
low.

Here is the output of dmesg with kernel 4.8:

dmesg | grep -i ath10k
[    9.264413] ath10k_pci 0000:02:00.0: pci irq msi oper_irq_mode 2 
irq_mode 0 reset_mode 0
[    9.544276] ath10k_pci 0000:02:00.0: Direct firmware load for 
ath10k/pre-cal-pci-0000:02:00.0.bin failed with error -2
[    9.544292] ath10k_pci 0000:02:00.0: Direct firmware load for 
ath10k/cal-pci-0000:02:00.0.bin failed with error -2
[    9.546101] ath10k_pci 0000:02:00.0: qca6174 hw2.1 target 0x05010000 
chip_id 0x003405ff sub 17aa:3044
[    9.546105] ath10k_pci 0000:02:00.0: kconfig debug 0 debugfs 0 
tracing 0 dfs 0 testmode 0
[    9.547303] ath10k_pci 0000:02:00.0: firmware ver 
SW_RM.1.1.1-00157-QCARMSWPZ-1 api 5 features ignore-otp,no-4addr-pad 
crc32 10bf8e08
[    9.613313] ath10k_pci 0000:02:00.0: board_file api 2 bmi_id N/A 
crc32 ae2e275a
[   10.813664] ath10k_pci 0000:02:00.0: htt-ver 3.1 wmi-op 4 htt-op 3 
cal otp max-sta 32 raw 0 hwcrypto 1
[   10.902302] ath10k_pci 0000:02:00.0 wlp2s0: renamed from wlan0
[   26.617451] ath10k_pci 0000:02:00.0: no channel configured; ignoring 
frame(s)!


Daniel Holz

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k