(no subject)

[Alexey Ivanov]

I'm trying to run OpenWrt(r48016) with ath10k driver on this device
(https://wikidevi.com/wiki/EnGenius_EAP1750H)

The calibration data is present at 0x5000@ART. I've copied it to board.bin

But the driver fails to load:
[   10.313982] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
irq_mode 0 reset_mode 0
[   10.526214] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/cal-pci-0000:00:00.0.bin failed with error -2
[   10.536740] ath10k_pci 0000:00:00.0: Falling back to user helper
[   10.611930] firmware ath10k!cal-pci-0000:00:00.0.bin:
firmware_loading_store: map pages failed
[   10.749685] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
0x4100016c chip_id 0x043202ff sub 0000:0000
[   10.759085] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
tracing 0 dfs 0 testmode 1
[   10.772154] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
features no-p2p crc32 f91e34f2
[   10.821280] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
[   10.831887] ath10k_pci 0000:00:00.0: Falling back to user helper
[   10.907421] firmware ath10k!QCA988X!hw2.0!board-2.bin:
firmware_loading_store: map pages failed
[   10.916658] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
crc32 bebc7c08
[   11.012769] ath10k_pci 0000:00:00.0: otp calibration failed: 2
[   11.018708] ath10k_pci 0000:00:00.0: failed to run otp: -22
[   11.024367] ath10k_pci 0000:00:00.0: could not init core (-22)
[   11.030370] ath10k_pci 0000:00:00.0: could not probe fw (-22)


The output after insmod ath10k_core skip_otp=Y:

[ 1346.161436] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
irq_mode 0 reset_mode 0
[ 1346.376382] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/cal-pci-0000:00:00.0.bin failed with error -2
[ 1346.386932] ath10k_pci 0000:00:00.0: Falling back to user helper
[ 1346.463861] firmware ath10k!cal-pci-0000:00:00.0.bin:
firmware_loading_store: map pages failed
[ 1346.475118] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
0x4100016c chip_id 0x043202ff sub 0000:0000
[ 1346.484521] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
tracing 0 dfs 0 testmode 1
[ 1346.497614] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
features no-p2p crc32 f91e34f2
[ 1346.548363] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
[ 1346.558999] ath10k_pci 0000:00:00.0: Falling back to user helper
[ 1346.635967] firmware ath10k!QCA988X!hw2.0!board-2.bin:
firmware_loading_store: map pages failed
[ 1346.645156] ath10k_pci 0000:00:00.0:
board_file api 1 bmi_id N/A crc32 bebc7c08
[ 1347.782528] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
cal otp max-sta 128 raw 0 hwcrypto 1


The original QSDK version by EnGenius works good.
What am I doing wrong? Or the
board_file api 1 is not supported by ath10k?

--
Best regards,
Alex Ivanov

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

Re:

[Michal Kazior]

On 13 January 2016 at 12:34, Alexey Ivanov <alexeyivan@gmail.com> wrote:
> I'm trying to run OpenWrt(r48016) with ath10k driver on this device
> (https://wikidevi.com/wiki/EnGenius_EAP1750H)
>
> The calibration data is present at 0x5000@ART. I've copied it to board.bin

Calibration data != board.bin.

You should put the data as cal-pci-0000:00:00.0.bin.


Michał

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

Re:

[Michal Kazior]

+ list

Please make sure to reply with the mailing list in CC next time.

On 13 January 2016 at 15:01, Alexey Ivanov <alexeyivan@gmail.com> wrote:
> Sorry for wrong subject
>
> after putting data to cal-pci-0000:00:00.0.bin:
> [   11.682188] firmware ath10k!cal-pci-0000:00:00.0.bin:
> firmware_loading_store: map pages failed
> other output is the same

Strange..


> Anyway, data at 0x5000 in ART looks like board.bin. It begins with
> 0x44 0x08 and contains string cus223-022-n1725 inside

Technically board.bin is more of a template. It doesn't have
calibration data and it doesn't have a mac address. These are
typically pulled from EEPROM of a given device by using otp.bin (it's
just a program that is executed on the device SoC/CPU).

When you consider most routers though their WLAN devices have the
EEPROM empty and have their calibration data stored out-of-band on
Flash partitions. These are basically board.bin files pre-filled with
mac address and calibration data, hence ath10k calls them "cal.bin".



Michał

> On 13 January 2016 at 17:12, Michal Kazior <michal.kazior@tieto.com> wrote:
>> On 13 January 2016 at 12:34, Alexey Ivanov <alexeyivan@gmail.com> wrote:
>>> I'm trying to run OpenWrt(r48016) with ath10k driver on this device
>>> (https://wikidevi.com/wiki/EnGenius_EAP1750H)
>>>
>>> The calibration data is present at 0x5000@ART. I've copied it to board.bin
>>
>> Calibration data != board.bin.
>>
>> You should put the data as cal-pci-0000:00:00.0.bin.
>>
>>
>> Michał
>
>
>
> --
> Best regards,
> Alex Ivanov

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

Re:

[Alexey Ivanov]

>Strange..

Do you have any idea what it can be?
"map pages failed" seems to be a result of failing to vmap() buffer in
fw_map_pages_buf() in firmware_loading_store()

-- 
Best regards,
Alex Ivanov

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

Re:

[Michal Kazior]

On 13 January 2016 at 15:45, Alexey Ivanov <alexeyivan@gmail.com> wrote:
>>Strange..
>
> Do you have any idea what it can be?
> "map pages failed" seems to be a result of failing to vmap() buffer in
> fw_map_pages_buf() in firmware_loading_store()

It's the same message as when you didn't have the cal.bin file at all.
Are you sure you placed it correctly? Perhaps a filename typo?

Another idea is initramfs which is used during early boot and which
doesn't include files from rootfs (meaning you'd have to rebuild it) -
not sure if this is the case though. You can rule it out by reloading
the driver after booting so that it surely has access to your rootfs's
/lib/firmware:

  rmmod ath10k_pci && modprobe ath10k_pci


Michał

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

Re:

[Alexey Ivanov]

Yes, you're right, Michał
putting cal.bin to lib/firmware/ath10k helped
(before I put it to lib/firmware/ath10k/QCA988X/hw2.0/)

but the mac address is still 00:03:7F:00:00:00

the output now is:
[   11.381198] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
irq_mode 0 reset_mode 0
[   11.725833] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
0x4100016c chip_id 0x043202ff sub 0000:0000
[   11.735226] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
tracing 0 dfs 0 testmode 1
[   11.748288] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
features no-p2p crc32 f91e34f2
[   11.797377] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
[   11.807985] ath10k_pci 0000:00:00.0: Falling back to user helper
[   11.882960] firmware ath10k!QCA988X!hw2.0!board-2.bin:
firmware_loading_store: map pages failed
[   11.893044] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
crc32 e623b3be
[   12.947352] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
cal file max-sta 128 raw 0 hwcrypto 1

As I understand, after correct calibration there should be normal mac?

On 13 January 2016 at 18:54, Michal Kazior <michal.kazior@tieto.com> wrote:
> On 13 January 2016 at 15:45, Alexey Ivanov <alexeyivan@gmail.com> wrote:
>>>Strange..
>>
>> Do you have any idea what it can be?
>> "map pages failed" seems to be a result of failing to vmap() buffer in
>> fw_map_pages_buf() in firmware_loading_store()
>
> It's the same message as when you didn't have the cal.bin file at all.
> Are you sure you placed it correctly? Perhaps a filename typo?
>
> Another idea is initramfs which is used during early boot and which
> doesn't include files from rootfs (meaning you'd have to rebuild it) -
> not sure if this is the case though. You can rule it out by reloading
> the driver after booting so that it surely has access to your rootfs's
> /lib/firmware:
>
>   rmmod ath10k_pci && modprobe ath10k_pci
>
>
> Michał



-- 
Best regards,
Alex Ivanov

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

Re:

[Michal Kazior]

On 14 January 2016 at 06:36, Alexey Ivanov <alexeyivan@gmail.com> wrote:
> Yes, you're right, Michał
> putting cal.bin to lib/firmware/ath10k helped
> (before I put it to lib/firmware/ath10k/QCA988X/hw2.0/)
>
> but the mac address is still 00:03:7F:00:00:00
>
> the output now is:
> [   11.381198] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
> irq_mode 0 reset_mode 0
> [   11.725833] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
> 0x4100016c chip_id 0x043202ff sub 0000:0000
> [   11.735226] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
> tracing 0 dfs 0 testmode 1
> [   11.748288] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
> features no-p2p crc32 f91e34f2
> [   11.797377] ath10k_pci 0000:00:00.0: Direct firmware load for
> ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
> [   11.807985] ath10k_pci 0000:00:00.0: Falling back to user helper
> [   11.882960] firmware ath10k!QCA988X!hw2.0!board-2.bin:
> firmware_loading_store: map pages failed
> [   11.893044] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
> crc32 e623b3be
> [   12.947352] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
> cal file max-sta 128 raw 0 hwcrypto 1
>
> As I understand, after correct calibration there should be normal mac?

Not necessarily. I think some routers have mac addresses elsewhere. If
you look at OpenWRT scripts you should probably find a few
scripts/hacks that put it in board.bin (the reason for the filename is
that at the time cal.bin support wasn't in ath10k yet).


Michał

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

Re:

[Alexey Ivanov]

OK, Michał
Thank you,

I found preinit scripts. some of them do ifconfig and some
do patching of the default mac in firmware.bin

On 14 January 2016 at 11:21, Michal Kazior <michal.kazior@tieto.com> wrote:
> On 14 January 2016 at 06:36, Alexey Ivanov <alexeyivan@gmail.com> wrote:
>> Yes, you're right, Michał
>> putting cal.bin to lib/firmware/ath10k helped
>> (before I put it to lib/firmware/ath10k/QCA988X/hw2.0/)
>>
>> but the mac address is still 00:03:7F:00:00:00
>>
>> the output now is:
>> [   11.381198] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
>> irq_mode 0 reset_mode 0
>> [   11.725833] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
>> 0x4100016c chip_id 0x043202ff sub 0000:0000
>> [   11.735226] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
>> tracing 0 dfs 0 testmode 1
>> [   11.748288] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
>> features no-p2p crc32 f91e34f2
>> [   11.797377] ath10k_pci 0000:00:00.0: Direct firmware load for
>> ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
>> [   11.807985] ath10k_pci 0000:00:00.0: Falling back to user helper
>> [   11.882960] firmware ath10k!QCA988X!hw2.0!board-2.bin:
>> firmware_loading_store: map pages failed
>> [   11.893044] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
>> crc32 e623b3be
>> [   12.947352] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
>> cal file max-sta 128 raw 0 hwcrypto 1
>>
>> As I understand, after correct calibration there should be normal mac?
>
> Not necessarily. I think some routers have mac addresses elsewhere. If
> you look at OpenWRT scripts you should probably find a few
> scripts/hacks that put it in board.bin (the reason for the filename is
> that at the time cal.bin support wasn't in ath10k yet).
>
>
> Michał



-- 
Best regards,
Alex Ivanov

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

Re: Re:

[Shajakhan, Mohammed Shafi (Mohammed Shafi)]

Hi Alex,

not sure this thread might be useful to you
https://lists.openwrt.org/pipermail/openwrt-devel/2015-June/033894.html

if the script is part of preinit and create .bin files, it will executed before wifi comes up.

regards
shafi
________________________________________
From: ath10k <ath10k-bounces@lists.infradead.org> on behalf of Alexey Ivanov <alexeyivan@gmail.com>
Sent: 14 January 2016 16:44
To: michal.kazior@tieto.com
Cc: ath10k@lists.infradead.org
Subject: Re:

OK, Michał
Thank you,

I found preinit scripts. some of them do ifconfig and some
do patching of the default mac in firmware.bin

On 14 January 2016 at 11:21, Michal Kazior <michal.kazior@tieto.com> wrote:
> On 14 January 2016 at 06:36, Alexey Ivanov <alexeyivan@gmail.com> wrote:
>> Yes, you're right, Michał
>> putting cal.bin to lib/firmware/ath10k helped
>> (before I put it to lib/firmware/ath10k/QCA988X/hw2.0/)
>>
>> but the mac address is still 00:03:7F:00:00:00
>>
>> the output now is:
>> [   11.381198] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
>> irq_mode 0 reset_mode 0
>> [   11.725833] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
>> 0x4100016c chip_id 0x043202ff sub 0000:0000
>> [   11.735226] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
>> tracing 0 dfs 0 testmode 1
>> [   11.748288] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
>> features no-p2p crc32 f91e34f2
>> [   11.797377] ath10k_pci 0000:00:00.0: Direct firmware load for
>> ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
>> [   11.807985] ath10k_pci 0000:00:00.0: Falling back to user helper
>> [   11.882960] firmware ath10k!QCA988X!hw2.0!board-2.bin:
>> firmware_loading_store: map pages failed
>> [   11.893044] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
>> crc32 e623b3be
>> [   12.947352] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
>> cal file max-sta 128 raw 0 hwcrypto 1
>>
>> As I understand, after correct calibration there should be normal mac?
>
> Not necessarily. I think some routers have mac addresses elsewhere. If
> you look at OpenWRT scripts you should probably find a few
> scripts/hacks that put it in board.bin (the reason for the filename is
> that at the time cal.bin support wasn't in ath10k yet).
>
>
> Michał



--
Best regards,
Alex Ivanov

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

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

Re: Re:

[Alexey Ivanov]

Thank you, Mohammed Shafi,

I'll try to impement it

On 14 January 2016 at 15:26, Shajakhan, Mohammed Shafi (Mohammed
Shafi) <mohammed@qti.qualcomm.com> wrote:
> Hi Alex,
>
> not sure this thread might be useful to you
> https://lists.openwrt.org/pipermail/openwrt-devel/2015-June/033894.html
>
> if the script is part of preinit and create .bin files, it will executed before wifi comes up.
>
> regards
> shafi
> ________________________________________
> From: ath10k <ath10k-bounces@lists.infradead.org> on behalf of Alexey Ivanov <alexeyivan@gmail.com>
> Sent: 14 January 2016 16:44
> To: michal.kazior@tieto.com
> Cc: ath10k@lists.infradead.org
> Subject: Re:
>
> OK, Michał
> Thank you,
>
> I found preinit scripts. some of them do ifconfig and some
> do patching of the default mac in firmware.bin
>
> On 14 January 2016 at 11:21, Michal Kazior <michal.kazior@tieto.com> wrote:
>> On 14 January 2016 at 06:36, Alexey Ivanov <alexeyivan@gmail.com> wrote:
>>> Yes, you're right, Michał
>>> putting cal.bin to lib/firmware/ath10k helped
>>> (before I put it to lib/firmware/ath10k/QCA988X/hw2.0/)
>>>
>>> but the mac address is still 00:03:7F:00:00:00
>>>
>>> the output now is:
>>> [   11.381198] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
>>> irq_mode 0 reset_mode 0
>>> [   11.725833] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
>>> 0x4100016c chip_id 0x043202ff sub 0000:0000
>>> [   11.735226] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
>>> tracing 0 dfs 0 testmode 1
>>> [   11.748288] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
>>> features no-p2p crc32 f91e34f2
>>> [   11.797377] ath10k_pci 0000:00:00.0: Direct firmware load for
>>> ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
>>> [   11.807985] ath10k_pci 0000:00:00.0: Falling back to user helper
>>> [   11.882960] firmware ath10k!QCA988X!hw2.0!board-2.bin:
>>> firmware_loading_store: map pages failed
>>> [   11.893044] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
>>> crc32 e623b3be
>>> [   12.947352] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
>>> cal file max-sta 128 raw 0 hwcrypto 1
>>>
>>> As I understand, after correct calibration there should be normal mac?
>>
>> Not necessarily. I think some routers have mac addresses elsewhere. If
>> you look at OpenWRT scripts you should probably find a few
>> scripts/hacks that put it in board.bin (the reason for the filename is
>> that at the time cal.bin support wasn't in ath10k yet).
>>
>>
>> Michał
>
>
>
> --
> Best regards,
> Alex Ivanov
>
> _______________________________________________
> ath10k mailing list
> ath10k@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/ath10k



-- 
Best regards,
Alex Ivanov

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

Re:

[Peter Oh]

On 01/13/2016 09:36 PM, Alexey Ivanov wrote:
> Yes, you're right, Michał
> putting cal.bin to lib/firmware/ath10k helped
> (before I put it to lib/firmware/ath10k/QCA988X/hw2.0/)
>
> but the mac address is still 00:03:7F:00:00:00
cal file has a mac address in it and ath10k uses it as default unless 
some other scripts change it during boot.
> the output now is:
> [   11.381198] ath10k_pci 0000:00:00.0: pci irq legacy interrupts 0
> irq_mode 0 reset_mode 0
> [   11.725833] ath10k_pci 0000:00:00.0: qca988x hw2.0 target
> 0x4100016c chip_id 0x043202ff sub 0000:0000
> [   11.735226] ath10k_pci 0000:00:00.0: kconfig debug 1 debugfs 1
> tracing 0 dfs 0 testmode 1
> [   11.748288] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.97 api 5
> features no-p2p crc32 f91e34f2
> [   11.797377] ath10k_pci 0000:00:00.0: Direct firmware load for
> ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
> [   11.807985] ath10k_pci 0000:00:00.0: Falling back to user helper
> [   11.882960] firmware ath10k!QCA988X!hw2.0!board-2.bin:
> firmware_loading_store: map pages failed
> [   11.893044] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
> crc32 e623b3be
> [   12.947352] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
> cal file max-sta 128 raw 0 hwcrypto 1
>
> As I understand, after correct calibration there should be normal mac?
>
> On 13 January 2016 at 18:54, Michal Kazior <michal.kazior@tieto.com> wrote:
>> On 13 January 2016 at 15:45, Alexey Ivanov <alexeyivan@gmail.com> wrote:
>>>> Strange..
>>> Do you have any idea what it can be?
>>> "map pages failed" seems to be a result of failing to vmap() buffer in
>>> fw_map_pages_buf() in firmware_loading_store()
>> It's the same message as when you didn't have the cal.bin file at all.
>> Are you sure you placed it correctly? Perhaps a filename typo?
>>
>> Another idea is initramfs which is used during early boot and which
>> doesn't include files from rootfs (meaning you'd have to rebuild it) -
>> not sure if this is the case though. You can rule it out by reloading
>> the driver after booting so that it surely has access to your rootfs's
>> /lib/firmware:
>>
>>    rmmod ath10k_pci && modprobe ath10k_pci
>>
>>
>> Michał
>
>


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