2022-11-10 07:08:29

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/9/22 19:35, Andy Green wrote:
>
> On 11/9/22 18:26, Arend Van Spriel wrote:

>> Can you share any kernel logs preferably with debug=0x1416 passed
>> during insmod.
>>
> I passed that in modprobe.d/ but the only logs are its usual ones and
> the ones I added

It seems the debug logging needs a few additional config options
additionally.

[ 3.469000] brcmfmac: brcmfmac_module_init No platform data available.
[ 3.469140] brcmfmac 0002:21:00.0: enabling device (0000 -> 0002)
[ 3.469179] brcmfmac: brcmf_chip_recognition found AXI chip: BCM43752/2
[ 3.470077] brcmfmac: brcmf_chip_cores_check [1 ] core 0x800:65
base 0x18000000 wrap 0x18100000
[ 3.470079] brcmfmac: brcmf_chip_cores_check [2 ] core 0x812:82
base 0x18001000 wrap 0x18101000
[ 3.470081] brcmfmac: brcmf_chip_cores_check [3 ] core 0x83e:16
base 0x18002000 wrap 0x18102000
[ 3.470082] brcmfmac: brcmf_chip_cores_check [4 ] core 0x83c:65
base 0x18003000 wrap 0x18103000
[ 3.470084] brcmfmac: brcmf_chip_cores_check [5 ] core 0x829:26
base 0x18004000 wrap 0x18104000
[ 3.470085] brcmfmac: brcmf_chip_cores_check [6 ] core 0x844:2
base 0x18005000 wrap 0x18105000
[ 3.470087] brcmfmac: brcmf_chip_cores_check [7 ] core 0x840:15
base 0x18010000 wrap 0x00000000
[ 3.470088] brcmfmac: brcmf_chip_cores_check [8 ] core 0x827:35
base 0x18012000 wrap 0x00000000
[ 3.470090] brcmfmac: brcmf_chip_cores_check [9 ] core 0x240:0
base 0x00000000 wrap 0x1810a000
[ 3.470091] brcmfmac: brcmf_chip_cores_check [10] core 0x135:0
base 0x00000000 wrap 0x18112000
[ 3.470092] brcmfmac: brcmf_chip_cores_check [11] core 0x135:0
base 0x00000000 wrap 0x18113000
[ 3.470094] brcmfmac: brcmf_chip_set_passive Enter
[ 3.579150] brcmfmac: brcmf_chip_set_passive Enter
[ 3.579630] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.579631] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.579655] brcmfmac: brcmf_chip_setup ccrev=65, pmurev=35,
pmucaps=0x84565f23
[ 3.579657] brcmfmac: brcmf_get_module_param Enter, bus=2,
chip=43752, rev=2
[ 3.579663] brcmfmac: brcmf_alloc Enter
[ 3.579675] brcmfmac: brcmf_fw_alloc_request: using
brcm/fw_bcm43752a2_pcie for chip BCM43752/2
[ 3.579678] brcmfmac: brcmf_fw_get_firmwares enter: dev=0002:21:00.0
[ 3.579680] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin
[ 3.583510] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.bin found
[ 3.583516] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.txt
[ 3.583726] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.txt found
[ 3.583727] brcmfmac: brcmf_fw_request_nvram_done enter: dev=0002:21:00.0
[ 3.583782] brcmfmac: brcmf_fw_request_nvram_done nvram
000000001685ccfe len 6040
[ 3.583784] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.clm_blob
[ 3.584081] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.clm_blob found
[ 3.584278] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.584279] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.677816] brcmfmac: Download NVRAM brcm/fw_bcm43752a2_pcie.txt
[ 3.678423] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
0xfa1a05e5
[ 3.678425] brcmfmac: brcmf_chip_set_active Enter
[ 9.707127] brcmfmac 0002:21:00.0: brcmf_pcie_download_fw_nvram: FW
failed to initialize
[ 9.707144] brcmfmac: brcmf_detach Enter
[ 9.731123] brcmfmac: brcmf_bus_change_state 0 -> 0
[ 9.731126] brcmfmac: brcmf_proto_detach Enter

-Andy


2022-11-10 07:36:03

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On November 10, 2022 8:06:38 AM Andy Green <[email protected]> wrote:

> On 11/9/22 19:35, Andy Green wrote:
>>
>> On 11/9/22 18:26, Arend Van Spriel wrote:
>
>>> Can you share any kernel logs preferably with debug=0x1416 passed
>>> during insmod.
>> I passed that in modprobe.d/ but the only logs are its usual ones and
>> the ones I added
>
> It seems the debug logging needs a few additional config options
> additionally.

Right. I wasn't sure you were aware ;-)

Thanks,
Arend



Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-13 08:47:54

by Shengyu Qu

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

Hello, I'm facing the same problem on sdio version of AP6275 module in mainline kernel. But I found a very old version from here could work with mainline brcmfmac driver.
Link: https://bbs.t-firefly.com/forum.php?mod=viewthread&tid=3107
Besides who I should ask to make proper firmware for it into linux-firmware repo?

Shengyu

2022-11-13 09:19:55

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/13/22 08:26, Shengyu Qu wrote:

> Hello, I'm facing the same problem on sdio version of AP6275 module in mainline kernel. But I found a very old version from here could work with mainline brcmfmac driver.
> Link: https://bbs.t-firefly.com/forum.php?mod=viewthread&tid=3107

Thanks for the info.

The firmware on the firefly thread doesn't seem to have a CLM blob that
the mainline driver wants, just two types of the ~900KB main firmware
for AP mode support or not.

It also doesn't have _sdio or _pcie in the filenames... I tried it but
although it's different in what it read after downloading NVRAM

[ 3.775602] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
0xfa2c05d3

(the rockchip SDK firmware says 0xfa1a05e5)

it still times it out a few seconds later as unresponsive after the fw
load... maybe it is for SDIO only, or the lack of CLM, present in the
Rockchip fw, means it won't get anywhere.

> Besides who I should ask to make proper firmware for it into linux-firmware repo?

It would also be great to hear if the PCIe firmware from Rockchip aimed
at the OOT driver was definitively incompatible with the mainline driver
and we have to find a solution for that from Broadcom, or it is expected
to basically work. I realize it might not be easy to get an answer
internally but this chipset is likely to appear in more RK3588 SBCs soon
since it's the one on the evb.

-Andy

> Shengyu

2022-11-13 09:25:17

by Shengyu Qu

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

Actually you can use any other clm info blob with this firmware as I tried. I also uploaded this to armbian's firmware repo. Hope we could get the proper firmware uploaded by Broadcom soon.

Shengyu

于 2022年11月13日 GMT+08:00 下午5:03:06, Andy Green <[email protected]> 写到:
>
>
>On 11/13/22 08:26, Shengyu Qu wrote:
>
>> Hello, I'm facing the same problem on sdio version of AP6275 module in mainline kernel. But I found a very old version from here could work with mainline brcmfmac driver.
>> Link: https://bbs.t-firefly.com/forum.php?mod=viewthread&tid=3107
>
>Thanks for the info.
>
>The firmware on the firefly thread doesn't seem to have a CLM blob that the mainline driver wants, just two types of the ~900KB main firmware for AP mode support or not.
>
>It also doesn't have _sdio or _pcie in the filenames... I tried it but although it's different in what it read after downloading NVRAM
>
>[ 3.775602] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr 0xfa2c05d3
>
>(the rockchip SDK firmware says 0xfa1a05e5)
>
>it still times it out a few seconds later as unresponsive after the fw load... maybe it is for SDIO only, or the lack of CLM, present in the Rockchip fw, means it won't get anywhere.
>
>> Besides who I should ask to make proper firmware for it into linux-firmware repo?
>
>It would also be great to hear if the PCIe firmware from Rockchip aimed at the OOT driver was definitively incompatible with the mainline driver and we have to find a solution for that from Broadcom, or it is expected to basically work. I realize it might not be easy to get an answer internally but this chipset is likely to appear in more RK3588 SBCs soon since it's the one on the evb.
>
>-Andy
>
>> Shengyu

2022-11-13 09:54:33

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/13/22 09:10, Shengyu Qu wrote:
> Actually you can use any other clm info blob with this firmware as I tried. I also uploaded this to armbian's firmware repo. Hope we could get the proper firmware uploaded by Broadcom soon.

Thank, I tried this, it did accept to download the clm part from the
Rockchip SDK firmware.

# ls -l /lib/firmware/brcm
-rw-r--r-- 1 root root 30993 Oct 3 08:40 clm_bcm43752a2_pcie_ag.blob
-rw-r--r-- 1 root root 743607 Nov 13 08:45 fw_bcm43752a2_ag.bin
lrwxrwxrwx 1 root root 46 Nov 13 09:19 fw_bcm43752a2_pcie.clm_blob
-> /lib/firmware/brcm/clm_bcm43752a2_pcie_ag.blob
lrwxrwxrwx 1 root root 20 Nov 13 08:38
fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin -> fw_bcm43752a2_ag.bin
lrwxrwxrwx 1 root root 36 Nov 13 08:47 fw_bcm43752a2_pcie.txt ->
/lib/firmware/brcm/nvram_ap6275s.txt
-rw-r--r-- 1 root root 8709 Nov 13 08:44 nvram_ap6275s.txt

But it's no happier after the NVRAM download part.

# dmesg | grep brcm
[ 3.530116] brcmfmac: brcmfmac_module_init No platform data available.
[ 3.530303] brcmfmac 0002:21:00.0: enabling device (0000 -> 0002)
[ 3.530362] brcmfmac: brcmf_chip_recognition found AXI chip: BCM43752/2
[ 3.531325] brcmfmac: brcmf_chip_cores_check [1 ] core 0x800:65
base 0x18000000 wrap 0x18100000
[ 3.531329] brcmfmac: brcmf_chip_cores_check [2 ] core 0x812:82
base 0x18001000 wrap 0x18101000
[ 3.531334] brcmfmac: brcmf_chip_cores_check [3 ] core 0x83e:16
base 0x18002000 wrap 0x18102000
[ 3.531338] brcmfmac: brcmf_chip_cores_check [4 ] core 0x83c:65
base 0x18003000 wrap 0x18103000
[ 3.531342] brcmfmac: brcmf_chip_cores_check [5 ] core 0x829:26
base 0x18004000 wrap 0x18104000
[ 3.531346] brcmfmac: brcmf_chip_cores_check [6 ] core 0x844:2
base 0x18005000 wrap 0x18105000
[ 3.531350] brcmfmac: brcmf_chip_cores_check [7 ] core 0x840:15
base 0x18010000 wrap 0x00000000
[ 3.531354] brcmfmac: brcmf_chip_cores_check [8 ] core 0x827:35
base 0x18012000 wrap 0x00000000
[ 3.531358] brcmfmac: brcmf_chip_cores_check [9 ] core 0x240:0
base 0x00000000 wrap 0x1810a000
[ 3.531362] brcmfmac: brcmf_chip_cores_check [10] core 0x135:0
base 0x00000000 wrap 0x18112000
[ 3.531365] brcmfmac: brcmf_chip_cores_check [11] core 0x135:0
base 0x00000000 wrap 0x18113000
[ 3.531369] brcmfmac: brcmf_chip_set_passive Enter
[ 3.641635] brcmfmac: brcmf_chip_set_passive Enter
[ 3.642156] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.642159] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.642187] brcmfmac: brcmf_chip_setup ccrev=65, pmurev=35,
pmucaps=0x84565f23
[ 3.642193] brcmfmac: brcmf_get_module_param Enter, bus=2,
chip=43752, rev=2
[ 3.642207] brcmfmac: brcmf_alloc Enter
[ 3.642231] brcmfmac: brcmf_fw_alloc_request: using
brcm/fw_bcm43752a2_pcie for chip BCM43752/2
[ 3.642235] brcmfmac: brcmf_fw_get_firmwares enter: dev=0002:21:00.0
[ 3.642241] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin
[ 3.646995] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.bin found
[ 3.647006] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.txt
[ 3.647308] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.txt found
[ 3.647312] brcmfmac: brcmf_fw_request_nvram_done enter: dev=0002:21:00.0
[ 3.647484] brcmfmac: brcmf_fw_request_nvram_done nvram
00000000c4269a30 len 5968
[ 3.647491] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.clm_blob
[ 3.647911] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.clm_blob found
[ 3.648197] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.648200] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.723260] brcmfmac: Download NVRAM brcm/fw_bcm43752a2_pcie.txt
[ 3.723895] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
0xfa2c05d3
[ 3.723899] brcmfmac: brcmf_chip_set_active Enter
[ 9.721628] brcmfmac 0002:21:00.0: brcmf_pcie_download_fw_nvram: FW
failed to initialize: 0xfa2c05d3
[ 9.721657] brcmfmac: brcmf_detach Enter
[ 9.745615] brcmfmac: brcmf_bus_change_state 0 -> 0
[ 9.745623] brcmfmac: brcmf_proto_detach Enter

CONFIG_BRCMUTIL=m
# CONFIG_BRCMSMAC is not set
CONFIG_BRCMFMAC=m
# CONFIG_BRCMFMAC_PROTO_MSGBUF is not set
# CONFIG_BRCMFMAC_SDIO is not set
# CONFIG_BRCMFMAC_USB is not set
CONFIG_BRCMFMAC_PCIE=y
CONFIG_BRCM_TRACING=y
CONFIG_BRCMDBG=y

I have the AP6275 module EN signal with the gpio wired to it as an
"always-on" gpio regulator in DT, and don't use the hacky rfkill thing
the rockchip SDK came with.

It's on top of 6.1.0-rc4.

Does your setup differ from this?

-Andy

> Shengyu
>
> 于 2022年11月13日 GMT+08:00 下午5:03:06, Andy Green <[email protected]> 写到:
>>
>>
>> On 11/13/22 08:26, Shengyu Qu wrote:
>>
>>> Hello, I'm facing the same problem on sdio version of AP6275 module in mainline kernel. But I found a very old version from here could work with mainline brcmfmac driver.
>>> Link: https://bbs.t-firefly.com/forum.php?mod=viewthread&tid=3107
>>
>> Thanks for the info.
>>
>> The firmware on the firefly thread doesn't seem to have a CLM blob that the mainline driver wants, just two types of the ~900KB main firmware for AP mode support or not.
>>
>> It also doesn't have _sdio or _pcie in the filenames... I tried it but although it's different in what it read after downloading NVRAM
>>
>> [ 3.775602] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr 0xfa2c05d3
>>
>> (the rockchip SDK firmware says 0xfa1a05e5)
>>
>> it still times it out a few seconds later as unresponsive after the fw load... maybe it is for SDIO only, or the lack of CLM, present in the Rockchip fw, means it won't get anywhere.
>>
>>> Besides who I should ask to make proper firmware for it into linux-firmware repo?
>>
>> It would also be great to hear if the PCIe firmware from Rockchip aimed at the OOT driver was definitively incompatible with the mainline driver and we have to find a solution for that from Broadcom, or it is expected to basically work. I realize it might not be easy to get an answer internally but this chipset is likely to appear in more RK3588 SBCs soon since it's the one on the evb.
>>
>> -Andy
>>
>>> Shengyu

2022-11-13 10:57:21

by Shengyu Qu

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

Almost same, except I'm using sdio driver (radxa rock 3a+their a6 module (AP6275S)). And I'm using mainline 5.19.8(self-compiled armbian edge) while I was testing.

Shengyu

于 2022年11月13日 GMT+08:00 下午5:41:19, Andy Green <[email protected]> 写到:
>
>
>On 11/13/22 09:10, Shengyu Qu wrote:
>> Actually you can use any other clm info blob with this firmware as I tried. I also uploaded this to armbian's firmware repo. Hope we could get the proper firmware uploaded by Broadcom soon.
>
>Thank, I tried this, it did accept to download the clm part from the Rockchip SDK firmware.
>
># ls -l /lib/firmware/brcm
>-rw-r--r-- 1 root root 30993 Oct 3 08:40 clm_bcm43752a2_pcie_ag.blob
>-rw-r--r-- 1 root root 743607 Nov 13 08:45 fw_bcm43752a2_ag.bin
>lrwxrwxrwx 1 root root 46 Nov 13 09:19 fw_bcm43752a2_pcie.clm_blob -> /lib/firmware/brcm/clm_bcm43752a2_pcie_ag.blob
>lrwxrwxrwx 1 root root 20 Nov 13 08:38 fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin -> fw_bcm43752a2_ag.bin
>lrwxrwxrwx 1 root root 36 Nov 13 08:47 fw_bcm43752a2_pcie.txt -> /lib/firmware/brcm/nvram_ap6275s.txt
>-rw-r--r-- 1 root root 8709 Nov 13 08:44 nvram_ap6275s.txt
>
>But it's no happier after the NVRAM download part.
>
># dmesg | grep brcm
>[ 3.530116] brcmfmac: brcmfmac_module_init No platform data available.
>[ 3.530303] brcmfmac 0002:21:00.0: enabling device (0000 -> 0002)
>[ 3.530362] brcmfmac: brcmf_chip_recognition found AXI chip: BCM43752/2
>[ 3.531325] brcmfmac: brcmf_chip_cores_check [1 ] core 0x800:65 base 0x18000000 wrap 0x18100000
>[ 3.531329] brcmfmac: brcmf_chip_cores_check [2 ] core 0x812:82 base 0x18001000 wrap 0x18101000
>[ 3.531334] brcmfmac: brcmf_chip_cores_check [3 ] core 0x83e:16 base 0x18002000 wrap 0x18102000
>[ 3.531338] brcmfmac: brcmf_chip_cores_check [4 ] core 0x83c:65 base 0x18003000 wrap 0x18103000
>[ 3.531342] brcmfmac: brcmf_chip_cores_check [5 ] core 0x829:26 base 0x18004000 wrap 0x18104000
>[ 3.531346] brcmfmac: brcmf_chip_cores_check [6 ] core 0x844:2 base 0x18005000 wrap 0x18105000
>[ 3.531350] brcmfmac: brcmf_chip_cores_check [7 ] core 0x840:15 base 0x18010000 wrap 0x00000000
>[ 3.531354] brcmfmac: brcmf_chip_cores_check [8 ] core 0x827:35 base 0x18012000 wrap 0x00000000
>[ 3.531358] brcmfmac: brcmf_chip_cores_check [9 ] core 0x240:0 base 0x00000000 wrap 0x1810a000
>[ 3.531362] brcmfmac: brcmf_chip_cores_check [10] core 0x135:0 base 0x00000000 wrap 0x18112000
>[ 3.531365] brcmfmac: brcmf_chip_cores_check [11] core 0x135:0 base 0x00000000 wrap 0x18113000
>[ 3.531369] brcmfmac: brcmf_chip_set_passive Enter
>[ 3.641635] brcmfmac: brcmf_chip_set_passive Enter
>[ 3.642156] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
>[ 3.642159] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000 size=1310720 (0x140000) sr=0 (0x0)
>[ 3.642187] brcmfmac: brcmf_chip_setup ccrev=65, pmurev=35, pmucaps=0x84565f23
>[ 3.642193] brcmfmac: brcmf_get_module_param Enter, bus=2, chip=43752, rev=2
>[ 3.642207] brcmfmac: brcmf_alloc Enter
>[ 3.642231] brcmfmac: brcmf_fw_alloc_request: using brcm/fw_bcm43752a2_pcie for chip BCM43752/2
>[ 3.642235] brcmfmac: brcmf_fw_get_firmwares enter: dev=0002:21:00.0
>[ 3.642241] brcmfmac: brcm_alt_fw_path FW alt path: brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin
>[ 3.646995] brcmfmac: brcmf_fw_complete_request firmware brcm/fw_bcm43752a2_pcie.bin found
>[ 3.647006] brcmfmac: brcm_alt_fw_path FW alt path: brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.txt
>[ 3.647308] brcmfmac: brcmf_fw_complete_request firmware brcm/fw_bcm43752a2_pcie.txt found
>[ 3.647312] brcmfmac: brcmf_fw_request_nvram_done enter: dev=0002:21:00.0
>[ 3.647484] brcmfmac: brcmf_fw_request_nvram_done nvram 00000000c4269a30 len 5968
>[ 3.647491] brcmfmac: brcm_alt_fw_path FW alt path: brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.clm_blob
>[ 3.647911] brcmfmac: brcmf_fw_complete_request firmware brcm/fw_bcm43752a2_pcie.clm_blob found
>[ 3.648197] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
>[ 3.648200] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000 size=1310720 (0x140000) sr=0 (0x0)
>[ 3.723260] brcmfmac: Download NVRAM brcm/fw_bcm43752a2_pcie.txt
>[ 3.723895] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr 0xfa2c05d3
>[ 3.723899] brcmfmac: brcmf_chip_set_active Enter
>[ 9.721628] brcmfmac 0002:21:00.0: brcmf_pcie_download_fw_nvram: FW failed to initialize: 0xfa2c05d3
>[ 9.721657] brcmfmac: brcmf_detach Enter
>[ 9.745615] brcmfmac: brcmf_bus_change_state 0 -> 0
>[ 9.745623] brcmfmac: brcmf_proto_detach Enter
>
>CONFIG_BRCMUTIL=m
># CONFIG_BRCMSMAC is not set
>CONFIG_BRCMFMAC=m
># CONFIG_BRCMFMAC_PROTO_MSGBUF is not set
># CONFIG_BRCMFMAC_SDIO is not set
># CONFIG_BRCMFMAC_USB is not set
>CONFIG_BRCMFMAC_PCIE=y
>CONFIG_BRCM_TRACING=y
>CONFIG_BRCMDBG=y
>
>I have the AP6275 module EN signal with the gpio wired to it as an "always-on" gpio regulator in DT, and don't use the hacky rfkill thing the rockchip SDK came with.
>
>It's on top of 6.1.0-rc4.
>
>Does your setup differ from this?
>
>-Andy
>
>> Shengyu
>>
>> 于 2022年11月13日 GMT+08:00 下午5:03:06, Andy Green <[email protected]> 写到:
>>>
>>>
>>> On 11/13/22 08:26, Shengyu Qu wrote:
>>>
>>>> Hello, I'm facing the same problem on sdio version of AP6275 module in mainline kernel. But I found a very old version from here could work with mainline brcmfmac driver.
>>>> Link: https://bbs.t-firefly.com/forum.php?mod=viewthread&tid=3107
>>>
>>> Thanks for the info.
>>>
>>> The firmware on the firefly thread doesn't seem to have a CLM blob that the mainline driver wants, just two types of the ~900KB main firmware for AP mode support or not.
>>>
>>> It also doesn't have _sdio or _pcie in the filenames... I tried it but although it's different in what it read after downloading NVRAM
>>>
>>> [ 3.775602] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr 0xfa2c05d3
>>>
>>> (the rockchip SDK firmware says 0xfa1a05e5)
>>>
>>> it still times it out a few seconds later as unresponsive after the fw load... maybe it is for SDIO only, or the lack of CLM, present in the Rockchip fw, means it won't get anywhere.
>>>
>>>> Besides who I should ask to make proper firmware for it into linux-firmware repo?
>>>
>>> It would also be great to hear if the PCIe firmware from Rockchip aimed at the OOT driver was definitively incompatible with the mainline driver and we have to find a solution for that from Broadcom, or it is expected to basically work. I realize it might not be easy to get an answer internally but this chipset is likely to appear in more RK3588 SBCs soon since it's the one on the evb.
>>>
>>> -Andy
>>>
>>>> Shengyu

2022-11-13 11:03:31

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On November 13, 2022 10:03:11 AM Andy Green <[email protected]> wrote:

> On 11/13/22 08:26, Shengyu Qu wrote:
>
>> Hello, I'm facing the same problem on sdio version of AP6275 module in
>> mainline kernel. But I found a very old version from here could work with
>> mainline brcmfmac driver.
>> Link: https://bbs.t-firefly.com/forum.php?mod=viewthread&tid=3107
>
> Thanks for the info.
>
> The firmware on the firefly thread doesn't seem to have a CLM blob that
> the mainline driver wants, just two types of the ~900KB main firmware
> for AP mode support or not.

A separate CLM blob may not be necessary when it is built-in in firmware
which can be overloaded by separate blob, but in that case it is optional.

> It also doesn't have _sdio or _pcie in the filenames... I tried it but
> although it's different in what it read after downloading NVRAM
>
> [ 3.775602] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
> 0xfa2c05d3
>
> (the rockchip SDK firmware says 0xfa1a05e5)
>
> it still times it out a few seconds later as unresponsive after the fw
> load... maybe it is for SDIO only, or the lack of CLM, present in the
> Rockchip fw, means it won't get anywhere.

That sharedram_addr looks wrong. Probably the firmware is not booting for
some reason. Can you provide a URL where I can download the rockchip
firmware? My internet search skills are lacking.

>
>> Besides who I should ask to make proper firmware for it into linux-firmware
>> repo?
>
> It would also be great to hear if the PCIe firmware from Rockchip aimed
> at the OOT driver was definitively incompatible with the mainline driver
> and we have to find a solution for that from Broadcom, or it is expected
> to basically work. I realize it might not be easy to get an answer
> internally but this chipset is likely to appear in more RK3588 SBCs soon
> since it's the one on the evb.

Right now I don't have enough information to comment on this.

Regards,
Arend



Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-13 16:02:32

by Shengyu Qu

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf


在 2022/11/13 18:56, Arend Van Spriel 写道:
> On November 13, 2022 10:03:11 AM Andy Green <[email protected]> wrote:
>
>> On 11/13/22 08:26, Shengyu Qu wrote:
>>
>>> Hello, I'm facing the same problem on sdio version of AP6275 module
>>> in mainline kernel. But I found a very old version from here could
>>> work with mainline brcmfmac driver.
>>> Link: https://bbs.t-firefly.com/forum.php?mod=viewthread&tid=3107
>>
>> Thanks for the info.
>>
>> The firmware on the firefly thread doesn't seem to have a CLM blob that
>> the mainline driver wants, just two types of the ~900KB main firmware
>> for AP mode support or not.
>
> A separate CLM blob may not be necessary when it is built-in in
> firmware which can be overloaded by separate blob, but in that case it
> is optional.
>
>> It also doesn't have _sdio or _pcie in the filenames... I tried it but
>> although it's different in what it read after downloading NVRAM
>>
>> [    3.775602] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
>> 0xfa2c05d3
>>
>> (the rockchip SDK firmware says 0xfa1a05e5)
>>
>> it still times it out a few seconds later as unresponsive after the fw
>> load... maybe it is for SDIO only, or the lack of CLM, present in the
>> Rockchip fw, means it won't get anywhere.
>
> That sharedram_addr looks wrong. Probably the firmware is not booting
> for some reason. Can you provide a URL where I can download the
> rockchip firmware? My internet search skills are lacking.
>
That error report happens when using sdio version firmware for pcie
card. If you want a new version of rockchip firmware, check this:
https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10/firmware/broadcom

Currently seems only firmware from firefly bbs above works with SDIO
version and mainline now. And no known version PCIE firmware can work
with mainline driver.

>>
>>> Besides who I should ask to make proper firmware for it into
>>> linux-firmware repo?
>>
>> It would also be great to hear if the PCIe firmware from Rockchip aimed
>> at the OOT driver was definitively incompatible with the mainline driver
>> and we have to find a solution for that from Broadcom, or it is expected
>> to basically work.  I realize it might not be easy to get an answer
>> internally but this chipset is likely to appear in more RK3588 SBCs soon
>> since it's the one on the evb.
>
> Right now I don't have enough information to comment on this.
Besides, could you upload the firmware that works with mainline driver
to linux-firmware repo?
>
> Regards,
> Arend
>
>

2022-11-13 17:44:55

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On November 13, 2022 4:52:45 PM Shengyu Qu <[email protected]> wrote:

> 在 2022/11/13 18:56, Arend Van Spriel 写道:

[snip]

>
>
>>>
>>>> Besides who I should ask to make proper firmware for it into
>>>> linux-firmware repo?
>>>
>>> It would also be great to hear if the PCIe firmware from Rockchip aimed
>>> at the OOT driver was definitively incompatible with the mainline driver
>>> and we have to find a solution for that from Broadcom, or it is expected
>>> to basically work. I realize it might not be easy to get an answer
>>> internally but this chipset is likely to appear in more RK3588 SBCs soon
>>> since it's the one on the evb.
>>
>> Right now I don't have enough information to comment on this.
> Besides, could you upload the firmware that works with mainline driver
> to linux-firmware repo?

I honestly don't know what entity provides the ampak wifi module. I guess
it is coming from Synaptics, but I am not sure. I rather fix the mainline
driver to make it work with the firmware (if possible) instead of releasing
a specific firmware for mainline.

Regards,
Arend




Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-13 18:48:47

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/13/22 17:35, Arend Van Spriel wrote:
> On November 13, 2022 4:52:45 PM Shengyu Qu <[email protected]> wrote:
>
>> 在 2022/11/13 18:56, Arend Van Spriel 写道:

>>>> to basically work.  I realize it might not be easy to get an answer
>>>> internally but this chipset is likely to appear in more RK3588 SBCs
>>>> soon
>>>> since it's the one on the evb.
>>>
>>> Right now I don't have enough information to comment on this.
>> Besides, could you upload the firmware that works with mainline driver
>> to linux-firmware repo?
>
> I honestly don't know what entity provides the ampak wifi module. I
> guess it is coming from Synaptics, but I am not sure. I rather fix the
> mainline driver to make it work with the firmware (if possible) instead
> of releasing a specific firmware for mainline.

That'd be great... I have some amount of time to help with this if
something useful I can do for pcie case.

-Andy


2022-11-16 12:29:28

by Shengyu Qu

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

Got a new version of downstream driver and firmware here: https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4

于 2022年11月14日 GMT+08:00 上午1:35:49, Arend Van Spriel <[email protected]> 写到:
>On November 13, 2022 4:52:45 PM Shengyu Qu <[email protected]> wrote:
>
>> 在 2022/11/13 18:56, Arend Van Spriel 写道:
>
>[snip]
>
>>
>>
>>>>
>>>>> Besides who I should ask to make proper firmware for it into
>>>>> linux-firmware repo?
>>>>
>>>> It would also be great to hear if the PCIe firmware from Rockchip aimed
>>>> at the OOT driver was definitively incompatible with the mainline driver
>>>> and we have to find a solution for that from Broadcom, or it is expected
>>>> to basically work. I realize it might not be easy to get an answer
>>>> internally but this chipset is likely to appear in more RK3588 SBCs soon
>>>> since it's the one on the evb.
>>>
>>> Right now I don't have enough information to comment on this.
>> Besides, could you upload the firmware that works with mainline driver
>> to linux-firmware repo?
>
>I honestly don't know what entity provides the ampak wifi module. I guess it is coming from Synaptics, but I am not sure. I rather fix the mainline driver to make it work with the firmware (if possible) instead of releasing a specific firmware for mainline.
>
>Regards,
>Arend
>
>
>

2022-11-16 14:30:18

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/16/22 12:25, Shengyu Qu wrote:

> Got a new version of downstream driver and firmware here: https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4

The AP6275 PCIE ones there are the firmware files from the rockchip Sept
SDK drop that I started with... they give the behaviour that the
sharedram address probed from the device is

[ 3.744635] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
0xfa1a05e5

and shortly after sending the nvram data

[ 9.876580] brcmfmac 0002:21:00.0: brcmf_pcie_download_fw_nvram: FW
failed to initialize: 0xfa1a05e5

-Andy


2022-11-17 09:28:41

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/16/2022 1:25 PM, Shengyu Qu wrote:
> Got a new version of downstream driver and firmware here: https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4

Thanks for that. The firmware indeed seems to originate from Synaptics
and this module can be found on wifi-alliance product finder:

https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275

We have no known contacts there and they do not seem to have much
interest in open-source. In that gitlab repo I see they are providing
bcmdhd driver and I assume they consider that sufficient.

Will see what I can do. First have to think on how to approach this
failure to get most out of a debug effort. I do not have the module over
here so we would need to go back and forth. The fact that the issue is
seen on both SDIO and PCIe is somewhat encouraging.

Regards,
Arend

> 于 2022年11月14日 GMT+08:00 上午1:35:49, Arend Van Spriel <[email protected]> 写到:
>> On November 13, 2022 4:52:45 PM Shengyu Qu <[email protected]> wrote:
>>
>>> 在 2022/11/13 18:56, Arend Van Spriel 写道:
>>
>> [snip]
>>
>>>
>>>
>>>>>
>>>>>> Besides who I should ask to make proper firmware for it into
>>>>>> linux-firmware repo?
>>>>>
>>>>> It would also be great to hear if the PCIe firmware from Rockchip aimed
>>>>> at the OOT driver was definitively incompatible with the mainline driver
>>>>> and we have to find a solution for that from Broadcom, or it is expected
>>>>> to basically work. I realize it might not be easy to get an answer
>>>>> internally but this chipset is likely to appear in more RK3588 SBCs soon
>>>>> since it's the one on the evb.
>>>>
>>>> Right now I don't have enough information to comment on this.
>>> Besides, could you upload the firmware that works with mainline driver
>>> to linux-firmware repo?
>>
>> I honestly don't know what entity provides the ampak wifi module. I guess it is coming from Synaptics, but I am not sure. I rather fix the mainline driver to make it work with the firmware (if possible) instead of releasing a specific firmware for mainline.
>>
>> Regards,
>> Arend
>>
>>
>>


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-17 12:02:34

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/17/22 09:23, Arend van Spriel wrote:
> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>> Got a new version of downstream driver and firmware here:
>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>
> Thanks for that. The firmware indeed seems to originate from Synaptics
> and this module can be found on wifi-alliance product finder:
>
> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>
> We have no known contacts there and they do not seem to have much
> interest in open-source. In that gitlab repo I see they are providing
> bcmdhd driver and I assume they consider that sufficient.
>
> Will see what I can do. First have to think on how to approach this
> failure to get most out of a debug effort. I do not have the module over
> here so we would need to go back and forth. The fact that the issue is
> seen on both SDIO and PCIe is somewhat encouraging.

Yes it seems that at some point with some firmware, SDIO has worked for,
eg, Angus. And, it should work well with bchdhd (but that does not seem
to work for scan for me on v6.1-rc5 uplevel with the firmware load
changed to use kernel apis; but the firmware has no timeout on bcmdhd).
So it doesn't feel a million miles away.

I have the hw, and (paid) time to help with debugging this "pair of
hands" style. I'm also a little familiar with the kernel code and can
try jiggle things autonomously if you can prompt what is interesting.

-Andy

> Regards,
> Arend

2022-11-17 16:18:04

by Angus Ainslie

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 2022-11-17 04:01, Andy Green wrote:
> On 11/17/22 09:23, Arend van Spriel wrote:
>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>> Got a new version of downstream driver and firmware here:
>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>
>> Thanks for that. The firmware indeed seems to originate from Synaptics
>> and this module can be found on wifi-alliance product finder:
>>
>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>
>> We have no known contacts there and they do not seem to have much
>> interest in open-source. In that gitlab repo I see they are providing
>> bcmdhd driver and I assume they consider that sufficient.
>>
>> Will see what I can do. First have to think on how to approach this
>> failure to get most out of a debug effort. I do not have the module
>> over here so we would need to go back and forth. The fact that the
>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>
> Yes it seems that at some point with some firmware, SDIO has worked
> for, eg, Angus. And, it should work well with bchdhd (but that does
> not seem to work for scan for me on v6.1-rc5 uplevel with the firmware
> load changed to use kernel apis; but the firmware has no timeout on
> bcmdhd). So it doesn't feel a million miles away.
>

This is the firmware we have been using for SDIO. It was sent to us
directly by SparkLAN so not sure how it compares to other variants.

https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm

Angus

> I have the hw, and (paid) time to help with debugging this "pair of
> hands" style. I'm also a little familiar with the kernel code and can
> try jiggle things autonomously if you can prompt what is interesting.
>
> -Andy
>
>> Regards,
>> Arend

2022-11-17 18:15:36

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]> wrote:

> On 2022-11-17 04:01, Andy Green wrote:
>> On 11/17/22 09:23, Arend van Spriel wrote:
>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>> Got a new version of downstream driver and firmware here:
>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>
>>> Thanks for that. The firmware indeed seems to originate from Synaptics
>>> and this module can be found on wifi-alliance product finder:
>>>
>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>
>>> We have no known contacts there and they do not seem to have much
>>> interest in open-source. In that gitlab repo I see they are providing
>>> bcmdhd driver and I assume they consider that sufficient.
>>>
>>> Will see what I can do. First have to think on how to approach this
>>> failure to get most out of a debug effort. I do not have the module
>>> over here so we would need to go back and forth. The fact that the
>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>
>> Yes it seems that at some point with some firmware, SDIO has worked
>> for, eg, Angus. And, it should work well with bchdhd (but that does
>> not seem to work for scan for me on v6.1-rc5 uplevel with the firmware
>> load changed to use kernel apis; but the firmware has no timeout on
>> bcmdhd). So it doesn't feel a million miles away.
>
> This is the firmware we have been using for SDIO. It was sent to us
> directly by SparkLAN so not sure how it compares to other variants.
>
> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm

It is a tad bit older:

$ strings brcmfmac43752-sdio.bin | grep FW-TAG
<FW-TAG>4362a2-roml/config_sdio_ampak_release_he Version=18.35.387.23.7
([email protected]) (wlan=r880297) Date=2020-07-13T10:34:00Z FWID=01-f50fdff0

Thanks,
Arend

>
> Angus
>
>> I have the hw, and (paid) time to help with debugging this "pair of
>> hands" style. I'm also a little familiar with the kernel code and can
>> try jiggle things autonomously if you can prompt what is interesting.
>>
>> -Andy
>>
>>> Regards,
>>> Arend




Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-17 18:38:07

by Angus Ainslie

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 2022-11-17 10:12, Arend Van Spriel wrote:
> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]> wrote:
>
>> On 2022-11-17 04:01, Andy Green wrote:
>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>> Got a new version of downstream driver and firmware here:
>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>
>>>> Thanks for that. The firmware indeed seems to originate from
>>>> Synaptics
>>>> and this module can be found on wifi-alliance product finder:
>>>>
>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>
>>>> We have no known contacts there and they do not seem to have much
>>>> interest in open-source. In that gitlab repo I see they are
>>>> providing
>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>
>>>> Will see what I can do. First have to think on how to approach this
>>>> failure to get most out of a debug effort. I do not have the module
>>>> over here so we would need to go back and forth. The fact that the
>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>
>>> Yes it seems that at some point with some firmware, SDIO has worked
>>> for, eg, Angus. And, it should work well with bchdhd (but that does
>>> not seem to work for scan for me on v6.1-rc5 uplevel with the
>>> firmware
>>> load changed to use kernel apis; but the firmware has no timeout on
>>> bcmdhd). So it doesn't feel a million miles away.
>>
>> This is the firmware we have been using for SDIO. It was sent to us
>> directly by SparkLAN so not sure how it compares to other variants.
>>
>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>
> It is a tad bit older:
>
> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>

Currently it's broken for hotspot mode but I haven't had the chance to
see if that is FW or something further up the stack.

Angus

> Thanks,
> Arend
>
>>
>> Angus
>>
>>> I have the hw, and (paid) time to help with debugging this "pair of
>>> hands" style. I'm also a little familiar with the kernel code and can
>>> try jiggle things autonomously if you can prompt what is interesting.
>>>
>>> -Andy
>>>
>>>> Regards,
>>>> Arend

2022-11-17 21:52:43

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On November 17, 2022 1:01:29 PM Andy Green <[email protected]> wrote:

> On 11/17/22 09:23, Arend van Spriel wrote:
>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>> Got a new version of downstream driver and firmware here:
>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>
>> Thanks for that. The firmware indeed seems to originate from Synaptics
>> and this module can be found on wifi-alliance product finder:
>>
>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>
>> We have no known contacts there and they do not seem to have much
>> interest in open-source. In that gitlab repo I see they are providing
>> bcmdhd driver and I assume they consider that sufficient.
>>
>> Will see what I can do. First have to think on how to approach this
>> failure to get most out of a debug effort. I do not have the module over
>> here so we would need to go back and forth. The fact that the issue is
>> seen on both SDIO and PCIe is somewhat encouraging.
>
> Yes it seems that at some point with some firmware, SDIO has worked for,
> eg, Angus. And, it should work well with bchdhd (but that does not seem
> to work for scan for me on v6.1-rc5 uplevel with the firmware load
> changed to use kernel apis; but the firmware has no timeout on bcmdhd).
> So it doesn't feel a million miles away.

I see. So could you load bcmdhd and provide log. Preferably with higher
debug level. It also has a module parameter for it. I think it is called
dhd_msg_level and please set it to 0x817.

Regards,
Arend



Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-18 05:49:31

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/17/22 21:52, Arend Van Spriel wrote:

> I see. So could you load bcmdhd and provide log. Preferably with higher
> debug level. It also has a module parameter for it. I think it is called
> dhd_msg_level and please set it to 0x817.

It's here:

https://warmcat.com/ap6275p-dhd-log1.txt

after boot he shows a wlan0

wlan0: flags=-28669<UP,BROADCAST,MULTICAST,DYNAMIC> mtu 1500
inet6 fe80::d69c:ddff:fef5:bf7e prefixlen 64 scopeid 0x20<link>
ether d4:9c:dd:f5:bf:7e txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 11 bytes 866 (866.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

The dhd OOT driver comes with a rockchip-hacked rfkill driver which has
all sorts of things in it related to wakelocks I removed, I built that
too and rfkill reports

# rfkill
ID TYPE DEVICE SOFT HARD
0 wlan phy0 blocked unblocked
1 wlan brcmfmac-wifi blocked unblocked

"brcmfmac-wifi" is a literal in the dhd driver sources, confusingly.
Mainline brcmfmac is not built or inserted as a module for this test.

# rfkill unblock

shows everything unblocked, but connmanctl can start scans but never get
any results. But it's happy with the firmware after hacking it to use
updated apis.

-Andy

2022-11-24 09:48:01

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/18/2022 6:41 AM, Andy Green wrote:
>
>
> On 11/17/22 21:52, Arend Van Spriel wrote:
>
>> I see. So could you load bcmdhd and provide log. Preferably with
>> higher debug level. It also has a module parameter for it. I think it
>> is called dhd_msg_level and please set it to 0x817.
>
> It's here:
>
> https://warmcat.com/ap6275p-dhd-log1.txt

Hi Andy,

Thanks for the log. Together with the dhd driver sources I could make
some sense of it. Perhaps you can try the following hack and see if we
make it over the first hurdle.

Regards,
Arend
---
diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c
b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c
index 97f0f13dfe50..ba436d71a7ec 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c
@@ -11,6 +11,7 @@
#include <linux/delay.h>
#include <linux/interrupt.h>
#include <linux/bcma/bcma.h>
+#include <linux/random.h>
#include <linux/sched.h>
#include <linux/io.h>
#include <asm/unaligned.h>
@@ -1527,6 +1528,33 @@ brcmf_pcie_init_share_ram_info(struct
brcmf_pciedev_info *devinfo,
return 0;
}

+#define BRCMF_SEED_SIGNATURE 0xFEEDC0DE
+#define BRCMF_ENTROPY_HOST_SEED_LEN 128
+
+struct brcmf_entropy_metadata {
+ __le32 count;
+ __le32 signature;
+};
+
+static void brcmf_pcie_write_entropy_bytes(struct brcmf_pciedev_info
*devinfo,
+ u32 nvram_len)
+{
+ struct brcmf_entropy_metadata entropy_info;
+ u8 seed[BRCMF_ENTROPY_HOST_SEED_LEN] = { 0 };
+ u32 count = sizeof(seed);
+ u32 addr = devinfo->ci->rambase + devinfo->ci->ramsize - nvram_len;
+
+ entropy_info.signture = cpu_to_le32(BRCMF_SEED_SIGNATURE);
+ entropy_info.count = cpu_to_le32(BRCMF_ENTROPY_HOST_SEED_LEN);
+
+ addr -= sizeof(entropy_info);
+ memcpy_toio(devinfo->tcm + addr, &entropy_info, sizeof(entropy_info));
+
+ get_random_bytes(seed, count);
+
+ addr -= sizeof(count);
+ memcpy_toio(devinfo->tcm + addr, seed, count);
+}

static int brcmf_pcie_download_fw_nvram(struct brcmf_pciedev_info
*devinfo,
const struct firmware *fw, void *nvram,
@@ -1568,6 +1596,8 @@ static int brcmf_pcie_download_fw_nvram(struct
brcmf_pciedev_info *devinfo,
devinfo->nvram_name);
}

+ brcmf_pcie_write_entropy_bytes(devinfo, nvram_len);
+
sharedram_addr_written = brcmf_pcie_read_ram32(devinfo,
devinfo->ci->ramsize -
4);


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-24 10:22:05

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/24/22 09:42, Arend van Spriel wrote:
> On 11/18/2022 6:41 AM, Andy Green wrote:
>>
>>
>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>
>>> I see. So could you load bcmdhd and provide log. Preferably with
>>> higher debug level. It also has a module parameter for it. I think it
>>> is called dhd_msg_level and please set it to 0x817.
>>
>> It's here:
>>
>> https://warmcat.com/ap6275p-dhd-log1.txt
>
> Hi Andy,
>
> Thanks for the log. Together with the dhd driver sources I could make
> some sense of it. Perhaps you can try the following hack and see if we
> make it over the first hurdle.

Thanks, I tried it, but although it looks necessary, there's evidently
something else since the result was the same.

> +
> +    entropy_info.signture = cpu_to_le32(BRCMF_SEED_SIGNATURE);

This should be ... .signature = ...

The debug logs

[ 3.487772] brcmfmac: brcmfmac_module_init No platform data available.
[ 3.487858] brcmfmac 0002:21:00.0: enabling device (0000 -> 0002)
[ 3.487896] brcmfmac: brcmf_chip_recognition found AXI chip: BCM43752/2
[ 3.488799] brcmfmac: brcmf_chip_cores_check [1 ] core 0x800:65
base 0x18000000 wrap 0x18100000
[ 3.488801] brcmfmac: brcmf_chip_cores_check [2 ] core 0x812:82
base 0x18001000 wrap 0x18101000
[ 3.488803] brcmfmac: brcmf_chip_cores_check [3 ] core 0x83e:16
base 0x18002000 wrap 0x18102000
[ 3.488804] brcmfmac: brcmf_chip_cores_check [4 ] core 0x83c:65
base 0x18003000 wrap 0x18103000
[ 3.488805] brcmfmac: brcmf_chip_cores_check [5 ] core 0x829:26
base 0x18004000 wrap 0x18104000
[ 3.488807] brcmfmac: brcmf_chip_cores_check [6 ] core 0x844:2
base 0x18005000 wrap 0x18105000
[ 3.488808] brcmfmac: brcmf_chip_cores_check [7 ] core 0x840:15
base 0x18010000 wrap 0x00000000
[ 3.488810] brcmfmac: brcmf_chip_cores_check [8 ] core 0x827:35
base 0x18012000 wrap 0x00000000
[ 3.488811] brcmfmac: brcmf_chip_cores_check [9 ] core 0x240:0
base 0x00000000 wrap 0x1810a000
[ 3.488813] brcmfmac: brcmf_chip_cores_check [10] core 0x135:0
base 0x00000000 wrap 0x18112000
[ 3.488814] brcmfmac: brcmf_chip_cores_check [11] core 0x135:0
base 0x00000000 wrap 0x18113000
[ 3.488815] brcmfmac: brcmf_chip_set_passive Enter
[ 3.594733] brcmfmac: brcmf_chip_set_passive Enter
[ 3.595214] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.595215] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.595238] brcmfmac: brcmf_chip_setup ccrev=65, pmurev=35,
pmucaps=0x84565f23
[ 3.595241] brcmfmac: brcmf_get_module_param Enter, bus=2,
chip=43752, rev=2
[ 3.595246] brcmfmac: brcmf_alloc Enter
[ 3.595254] brcmfmac: brcmf_fw_alloc_request: using
brcm/fw_bcm43752a2_pcie for chip BCM43752/2
[ 3.595257] brcmfmac: brcmf_fw_get_firmwares enter: dev=0002:21:00.0
[ 3.595259] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin
[ 3.595555] brcmfmac 0002:21:00.0: Direct firmware load for
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin failed with error -2
[ 3.600006] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.bin found
[ 3.600010] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.txt
[ 3.600323] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.txt found
[ 3.600324] brcmfmac: brcmf_fw_request_nvram_done enter: dev=0002:21:00.0
[ 3.600378] brcmfmac: brcmf_fw_request_nvram_done nvram
00000000eea1e0fd len 6040
[ 3.600381] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.clm_blob
[ 3.600641] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.clm_blob found
[ 3.600836] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.600837] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.694272] brcmfmac: Download NVRAM brcm/fw_bcm43752a2_pcie.txt
[ 3.694895] brcmfmac: brcmf_pcie_write_entropy_bytes: written 6040
entropy bytes
[ 3.694910] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
0xfa1a05e5
[ 3.694911] brcmfmac: brcmf_chip_set_active Enter
[ 9.774716] brcmfmac 0002:21:00.0: brcmf_pcie_download_fw_nvram: FW
failed to initialize: 0xfa1a05e5
[ 9.774733] brcmfmac: brcmf_detach Enter
[ 9.798714] brcmfmac: brcmf_bus_change_state 0 -> 0
[ 9.798716] brcmfmac: brcmf_proto_detach Enter

I added a log showing it sending the entropy bytes.

Firmware situation

# ls -l /lib/firmware/brcm/
total 956
-rw-r--r-- 1 root root 30993 Nov 16 12:31 clm_bcm43752a2_pcie_ag.blob
lrwxrwxrwx 1 root root 44 Nov 24 10:00 fw_bcm43752a2_pcie.bin ->
/lib/firmware/brcm/fw_bcm43752a2_pcie_ag.bin
lrwxrwxrwx 1 root root 46 Nov 24 10:02 fw_bcm43752a2_pcie.clm_blob
-> /lib/firmware/brcm/clm_bcm43752a2_pcie_ag.blob
lrwxrwxrwx 1 root root 36 Nov 24 10:02 fw_bcm43752a2_pcie.txt ->
/lib/firmware/brcm/nvram_AP6275P.txt
-rw-r--r-- 1 root root 936074 Nov 16 12:31 fw_bcm43752a2_pcie_ag.bin
-rw-r--r-- 1 root root 7458 Nov 16 12:31 nvram_AP6275P.txt

-Andy

2022-11-24 10:56:51

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/24/2022 11:20 AM, Andy Green wrote:
>
>
> On 11/24/22 09:42, Arend van Spriel wrote:
>> On 11/18/2022 6:41 AM, Andy Green wrote:
>>>
>>>
>>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>>
>>>> I see. So could you load bcmdhd and provide log. Preferably with
>>>> higher debug level. It also has a module parameter for it. I think
>>>> it is called dhd_msg_level and please set it to 0x817.
>>>
>>> It's here:
>>>
>>> https://warmcat.com/ap6275p-dhd-log1.txt
>>
>> Hi Andy,
>>
>> Thanks for the log. Together with the dhd driver sources I could make
>> some sense of it. Perhaps you can try the following hack and see if we
>> make it over the first hurdle.
>
> Thanks, I tried it, but although it looks necessary, there's evidently
> something else since the result was the same.
>
>> +
>> +    entropy_info.signture = cpu_to_le32(BRCMF_SEED_SIGNATURE);
>
> This should be ... .signature = ...

That's what you get if you get too confident... :-(

> The debug logs
>
> [    3.487772] brcmfmac: brcmfmac_module_init No platform data available.
> [    3.487858] brcmfmac 0002:21:00.0: enabling device (0000 -> 0002)
> [    3.487896] brcmfmac: brcmf_chip_recognition found AXI chip: BCM43752/2
> [    3.488799] brcmfmac: brcmf_chip_cores_check  [1 ] core 0x800:65 base
> 0x18000000 wrap 0x18100000
> [    3.488801] brcmfmac: brcmf_chip_cores_check  [2 ] core 0x812:82 base
> 0x18001000 wrap 0x18101000
> [    3.488803] brcmfmac: brcmf_chip_cores_check  [3 ] core 0x83e:16 base
> 0x18002000 wrap 0x18102000
> [    3.488804] brcmfmac: brcmf_chip_cores_check  [4 ] core 0x83c:65 base
> 0x18003000 wrap 0x18103000
> [    3.488805] brcmfmac: brcmf_chip_cores_check  [5 ] core 0x829:26 base
> 0x18004000 wrap 0x18104000
> [    3.488807] brcmfmac: brcmf_chip_cores_check  [6 ] core 0x844:2 base
> 0x18005000 wrap 0x18105000
> [    3.488808] brcmfmac: brcmf_chip_cores_check  [7 ] core 0x840:15 base
> 0x18010000 wrap 0x00000000
> [    3.488810] brcmfmac: brcmf_chip_cores_check  [8 ] core 0x827:35 base
> 0x18012000 wrap 0x00000000
> [    3.488811] brcmfmac: brcmf_chip_cores_check  [9 ] core 0x240:0 base
> 0x00000000 wrap 0x1810a000
> [    3.488813] brcmfmac: brcmf_chip_cores_check  [10] core 0x135:0 base
> 0x00000000 wrap 0x18112000
> [    3.488814] brcmfmac: brcmf_chip_cores_check  [11] core 0x135:0 base
> 0x00000000 wrap 0x18113000
> [    3.488815] brcmfmac: brcmf_chip_set_passive Enter
> [    3.594733] brcmfmac: brcmf_chip_set_passive Enter
> [    3.595214] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
> [    3.595215] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
> size=1310720 (0x140000) sr=0 (0x0)
> [    3.595238] brcmfmac: brcmf_chip_setup ccrev=65, pmurev=35,
> pmucaps=0x84565f23
> [    3.595241] brcmfmac: brcmf_get_module_param Enter, bus=2,
> chip=43752, rev=2
> [    3.595246] brcmfmac: brcmf_alloc Enter
> [    3.595254] brcmfmac: brcmf_fw_alloc_request: using
> brcm/fw_bcm43752a2_pcie for chip BCM43752/2
> [    3.595257] brcmfmac: brcmf_fw_get_firmwares enter: dev=0002:21:00.0
> [    3.595259] brcmfmac: brcm_alt_fw_path FW alt path:
> brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin
> [    3.595555] brcmfmac 0002:21:00.0: Direct firmware load for
> brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin failed with error -2
> [    3.600006] brcmfmac: brcmf_fw_complete_request firmware
> brcm/fw_bcm43752a2_pcie.bin found
> [    3.600010] brcmfmac: brcm_alt_fw_path FW alt path:
> brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.txt
> [    3.600323] brcmfmac: brcmf_fw_complete_request firmware
> brcm/fw_bcm43752a2_pcie.txt found
> [    3.600324] brcmfmac: brcmf_fw_request_nvram_done enter:
> dev=0002:21:00.0
> [    3.600378] brcmfmac: brcmf_fw_request_nvram_done nvram
> 00000000eea1e0fd len 6040
> [    3.600381] brcmfmac: brcm_alt_fw_path FW alt path:
> brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.clm_blob
> [    3.600641] brcmfmac: brcmf_fw_complete_request firmware
> brcm/fw_bcm43752a2_pcie.clm_blob found
> [    3.600836] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
> [    3.600837] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
> size=1310720 (0x140000) sr=0 (0x0)
> [    3.694272] brcmfmac: Download NVRAM brcm/fw_bcm43752a2_pcie.txt
> [    3.694895] brcmfmac: brcmf_pcie_write_entropy_bytes: written 6040
> entropy bytes
> [    3.694910] brcmfmac: brcmf_pcie_download_fw_nvram: sharedram_addr
> 0xfa1a05e5
> [    3.694911] brcmfmac: brcmf_chip_set_active Enter
> [    9.774716] brcmfmac 0002:21:00.0: brcmf_pcie_download_fw_nvram: FW
> failed to initialize: 0xfa1a05e5
> [    9.774733] brcmfmac: brcmf_detach Enter
> [    9.798714] brcmfmac: brcmf_bus_change_state 0 -> 0
> [    9.798716] brcmfmac: brcmf_proto_detach Enter
>
> I added a log showing it sending the entropy bytes.

Ok. Something is off there. Maybe I tried to be to smart. How does that
log statement look which you added?

Regards,
Arend


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-24 11:18:41

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/24/22 10:49, Arend van Spriel wrote:
> On 11/24/2022 11:20 AM, Andy Green wrote:

>>> Thanks for the log. Together with the dhd driver sources I could make
>>> some sense of it. Perhaps you can try the following hack and see if
>>> we make it over the first hurdle.
>>
>> Thanks, I tried it, but although it looks necessary, there's evidently
>> something else since the result was the same.
>>
>>> +
>>> +    entropy_info.signture = cpu_to_le32(BRCMF_SEED_SIGNATURE);
>>
>> This should be ... .signature = ...
>
> That's what you get if you get too confident... :-(

Just mention it in case anyone following along at home with the patches.

>> I added a log showing it sending the entropy bytes.
>
> Ok. Something is off there. Maybe I tried to be to smart. How does that
> log statement look which you added?

It's inline with the other logs, it seems to be coming before the actual
NVRAM download part

>> [ 3.694272] brcmfmac: Download NVRAM brcm/fw_bcm43752a2_pcie.txt
>> [ 3.694895] brcmfmac: brcmf_pcie_write_entropy_bytes: written 6040
>> entropy bytes

-Andy

2022-11-24 12:15:16

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/24/22 09:42, Arend van Spriel wrote:
> On 11/18/2022 6:41 AM, Andy Green wrote:
>>
>>
>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>
>>> I see. So could you load bcmdhd and provide log. Preferably with
>>> higher debug level. It also has a module parameter for it. I think it
>>> is called dhd_msg_level and please set it to 0x817.
>>
>> It's here:
>>
>> https://warmcat.com/ap6275p-dhd-log1.txt
>
> Hi Andy,
>
> Thanks for the log. Together with the dhd driver sources I could make
> some sense of it. Perhaps you can try the following hack and see if we
> make it over the first hurdle.

OH! I found that the line below should be count, not sizeof(count).

> +    addr -= sizeof(count);
> +    memcpy_toio(devinfo->tcm + addr, seed, count);
> +}

With that, you were right on the money, the firmware comes up

[ 3.796062] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM43752/2
wl0: Jul 12 2022 18:54:54 version 18.35.387.23.146 (g412cc5ec) FWID
01-93c53be6

along with

wlP2p33s0: flags=4098<BROADCAST,MULTICAST> mtu 1500
ether d4:9c:dd:f5:bf:7e txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

I will try get it to scan...

... any way, good job!

-Andy

2022-11-24 12:27:02

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/24/2022 1:14 PM, Andy Green wrote:
>
>
> On 11/24/22 09:42, Arend van Spriel wrote:
>> On 11/18/2022 6:41 AM, Andy Green wrote:
>>>
>>>
>>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>>
>>>> I see. So could you load bcmdhd and provide log. Preferably with
>>>> higher debug level. It also has a module parameter for it. I think
>>>> it is called dhd_msg_level and please set it to 0x817.
>>>
>>> It's here:
>>>
>>> https://warmcat.com/ap6275p-dhd-log1.txt
>>
>> Hi Andy,
>>
>> Thanks for the log. Together with the dhd driver sources I could make
>> some sense of it. Perhaps you can try the following hack and see if we
>> make it over the first hurdle.
>
> OH!  I found that the line below should be count, not sizeof(count).

Good job! I was testing you obviously. You passed. Nah, I just feel stupid.

>> +    addr -= sizeof(count);
>> +    memcpy_toio(devinfo->tcm + addr, seed, count);
>> +}
>
> With that, you were right on the money, the firmware comes up
>
> [    3.796062] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM43752/2
> wl0: Jul 12 2022 18:54:54 version 18.35.387.23.146 (g412cc5ec) FWID
> 01-93c53be6
>
> along with
>
> wlP2p33s0: flags=4098<BROADCAST,MULTICAST>  mtu 1500
>         ether d4:9c:dd:f5:bf:7e  txqueuelen 1000  (Ethernet)
>         RX packets 0  bytes 0 (0.0 B)
>         RX errors 0  dropped 0  overruns 0  frame 0
>         TX packets 0  bytes 0 (0.0 B)
>         TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>
> I will try get it to scan...
>
> ... any way, good job!

Thanks. Will clean it up and formally submit it. I also had a quick look
at the SDIO variant, but not seeing anything obvious there in terms of
firmware loading.

Regards,
Arend


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-24 12:27:21

by Shengyu Qu

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

Great work! If you need to test SDIO variant, I could help with it.
Shengyu

于 2022年11月24日 GMT+08:00 下午8:22:09, Arend van Spriel <[email protected]> 写到:
>On 11/24/2022 1:14 PM, Andy Green wrote:
>>
>>
>> On 11/24/22 09:42, Arend van Spriel wrote:
>>> On 11/18/2022 6:41 AM, Andy Green wrote:
>>>>
>>>>
>>>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>>>
>>>>> I see. So could you load bcmdhd and provide log. Preferably with higher debug level. It also has a module parameter for it. I think it is called dhd_msg_level and please set it to 0x817.
>>>>
>>>> It's here:
>>>>
>>>> https://warmcat.com/ap6275p-dhd-log1.txt
>>>
>>> Hi Andy,
>>>
>>> Thanks for the log. Together with the dhd driver sources I could make some sense of it. Perhaps you can try the following hack and see if we make it over the first hurdle.
>>
>> OH!  I found that the line below should be count, not sizeof(count).
>
>Good job! I was testing you obviously. You passed. Nah, I just feel stupid.
>
>>> +    addr -= sizeof(count);
>>> +    memcpy_toio(devinfo->tcm + addr, seed, count);
>>> +}
>>
>> With that, you were right on the money, the firmware comes up
>>
>> [    3.796062] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM43752/2 wl0: Jul 12 2022 18:54:54 version 18.35.387.23.146 (g412cc5ec) FWID 01-93c53be6
>>
>> along with
>>
>> wlP2p33s0: flags=4098<BROADCAST,MULTICAST>  mtu 1500
>>         ether d4:9c:dd:f5:bf:7e  txqueuelen 1000  (Ethernet)
>>         RX packets 0  bytes 0 (0.0 B)
>>         RX errors 0  dropped 0  overruns 0  frame 0
>>         TX packets 0  bytes 0 (0.0 B)
>>         TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>>
>> I will try get it to scan...
>>
>> ... any way, good job!
>
>Thanks. Will clean it up and formally submit it. I also had a quick look at the SDIO variant, but not seeing anything obvious there in terms of firmware loading.
>
>Regards,
>Arend

2022-11-24 12:27:41

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/24/22 12:22, Arend van Spriel wrote:
> On 11/24/2022 1:14 PM, Andy Green wrote:
>>
>>
>> On 11/24/22 09:42, Arend van Spriel wrote:
>>> On 11/18/2022 6:41 AM, Andy Green wrote:
>>>>
>>>>
>>>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>>>
>>>>> I see. So could you load bcmdhd and provide log. Preferably with
>>>>> higher debug level. It also has a module parameter for it. I think
>>>>> it is called dhd_msg_level and please set it to 0x817.
>>>>
>>>> It's here:
>>>>
>>>> https://warmcat.com/ap6275p-dhd-log1.txt
>>>
>>> Hi Andy,
>>>
>>> Thanks for the log. Together with the dhd driver sources I could make
>>> some sense of it. Perhaps you can try the following hack and see if
>>> we make it over the first hurdle.
>>
>> OH!  I found that the line below should be count, not sizeof(count).
>
> Good job! I was testing you obviously. You passed. Nah, I just feel stupid.

No it's great, the point is it directly identified the root cause
without access to the platform, it's not so easy.

>> ... any way, good job!
>
> Thanks. Will clean it up and formally submit it. I also had a quick look
> at the SDIO variant, but not seeing anything obvious there in terms of
> firmware loading.

I only have the pcie module not the SDIO one.

For the scan situation, he understands what is being asked via
connmanctl "enable wifi", but he feels something wrong

connmanctl> enable wifi

connmanctl>
Enabled wifi
connmanctl> [ 272.174946] ieee80211 phy0: brcmf_run_escan: error (-52)
[ 272.174956] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)


connmanctl> [ 275.133157] ieee80211 phy0: brcmf_run_escan: error (-52)
[ 275.133166] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)


connmanctl> [ 284.138852] ieee80211 phy0: brcmf_run_escan: error (-52)
[ 284.138856] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)

52 is EBADE, I recall you mentioned that the scan api had been changed
recently... is this possibly related to that? Grepping around for EBADE
near the function it mentions, it tries to take evasive action if it
sees one in some places already.

-Andy


2022-11-24 12:39:57

by Shengyu Qu

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

Seems SDIO variant load firmware correctly without any modification, but also reports this -52 error while scanning.
Shengyu

于 2022年11月24日 GMT+08:00 下午8:26:45, Andy Green <[email protected]> 写到:
>
>
>On 11/24/22 12:22, Arend van Spriel wrote:
>> On 11/24/2022 1:14 PM, Andy Green wrote:
>>>
>>>
>>> On 11/24/22 09:42, Arend van Spriel wrote:
>>>> On 11/18/2022 6:41 AM, Andy Green wrote:
>>>>>
>>>>>
>>>>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>>>>
>>>>>> I see. So could you load bcmdhd and provide log. Preferably with higher debug level. It also has a module parameter for it. I think it is called dhd_msg_level and please set it to 0x817.
>>>>>
>>>>> It's here:
>>>>>
>>>>> https://warmcat.com/ap6275p-dhd-log1.txt
>>>>
>>>> Hi Andy,
>>>>
>>>> Thanks for the log. Together with the dhd driver sources I could make some sense of it. Perhaps you can try the following hack and see if we make it over the first hurdle.
>>>
>>> OH!  I found that the line below should be count, not sizeof(count).
>>
>> Good job! I was testing you obviously. You passed. Nah, I just feel stupid.
>
>No it's great, the point is it directly identified the root cause without access to the platform, it's not so easy.
>
>>> ... any way, good job!
>>
>> Thanks. Will clean it up and formally submit it. I also had a quick look at the SDIO variant, but not seeing anything obvious there in terms of firmware loading.
>
>I only have the pcie module not the SDIO one.
>
>For the scan situation, he understands what is being asked via connmanctl "enable wifi", but he feels something wrong
>
>connmanctl> enable wifi
>
>connmanctl>
>Enabled wifi
>connmanctl> [ 272.174946] ieee80211 phy0: brcmf_run_escan: error (-52)
>[ 272.174956] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)
>
>
>connmanctl> [ 275.133157] ieee80211 phy0: brcmf_run_escan: error (-52)
>[ 275.133166] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)
>
>
>connmanctl> [ 284.138852] ieee80211 phy0: brcmf_run_escan: error (-52)
>[ 284.138856] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)
>
>52 is EBADE, I recall you mentioned that the scan api had been changed recently... is this possibly related to that? Grepping around for EBADE near the function it mentions, it tries to take evasive action if it sees one in some places already.
>
>-Andy
>
>

2022-11-24 12:40:16

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/24/2022 1:35 PM, Shengyu Qu wrote:
> Seems SDIO variant load firmware correctly without any modification, but also reports this -52 error while scanning.
> Shengyu

Thanks for letting me know. So we can focus on scanning. I will reply to
Andy about that.

Regards,
Arend


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-24 12:52:59

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/24/2022 1:26 PM, Andy Green wrote:
>
>
> On 11/24/22 12:22, Arend van Spriel wrote:
>> On 11/24/2022 1:14 PM, Andy Green wrote:
>>>
>>>
>>> On 11/24/22 09:42, Arend van Spriel wrote:
>>>> On 11/18/2022 6:41 AM, Andy Green wrote:
>>>>>
>>>>>
>>>>> On 11/17/22 21:52, Arend Van Spriel wrote:
>>>>>
>>>>>> I see. So could you load bcmdhd and provide log. Preferably with
>>>>>> higher debug level. It also has a module parameter for it. I think
>>>>>> it is called dhd_msg_level and please set it to 0x817.
>>>>>
>>>>> It's here:
>>>>>
>>>>> https://warmcat.com/ap6275p-dhd-log1.txt
>>>>
>>>> Hi Andy,
>>>>
>>>> Thanks for the log. Together with the dhd driver sources I could
>>>> make some sense of it. Perhaps you can try the following hack and
>>>> see if we make it over the first hurdle.
>>>
>>> OH!  I found that the line below should be count, not sizeof(count).
>>
>> Good job! I was testing you obviously. You passed. Nah, I just feel
>> stupid.
>
> No it's great, the point is it directly identified the root cause
> without access to the platform, it's not so easy.
>
>>> ... any way, good job!
>>
>> Thanks. Will clean it up and formally submit it. I also had a quick
>> look at the SDIO variant, but not seeing anything obvious there in
>> terms of firmware loading.
>
> I only have the pcie module not the SDIO one.
>
> For the scan situation, he understands what is being asked via
> connmanctl "enable wifi", but he feels something wrong
>
> connmanctl> enable wifi
>
> connmanctl>
> Enabled wifi
> connmanctl> [  272.174946] ieee80211 phy0: brcmf_run_escan: error (-52)
> [  272.174956] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)
>
>
> connmanctl> [  275.133157] ieee80211 phy0: brcmf_run_escan: error (-52)
> [  275.133166] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)
>
>
> connmanctl> [  284.138852] ieee80211 phy0: brcmf_run_escan: error (-52)
> [  284.138856] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)
>
> 52 is EBADE, I recall you mentioned that the scan api had been changed
> recently... is this possibly related to that?  Grepping around for EBADE
> near the function it mentions, it tries to take evasive action if it
> sees one in some places already.

EBADE is a generic error message indicating something went wrong
communicating with firmware. Not very informative. Were you using any
debug level? Anyway, please load the driver with module parameter
debug=0x5416 and try scanning using iw utility.

Regards,
Arend


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-24 12:57:25

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/24/22 12:44, Arend van Spriel wrote:

>> 52 is EBADE, I recall you mentioned that the scan api had been changed
>> recently... is this possibly related to that?  Grepping around for
>> EBADE near the function it mentions, it tries to take evasive action
>> if it sees one in some places already.
>
> EBADE is a generic error message indicating something went wrong
> communicating with firmware. Not very informative. Were you using any
> debug level? Anyway, please load the driver with module parameter
> debug=0x5416 and try scanning using iw utility.

My kernel crashed on boot while I was trying things earlier
necessitating a rootfs reflash, I lost the /etc/modprobe.d entry for the
debug.

Adding it back in, it's a lot more informative

[ 3.569646] brcmfmac: brcmfmac_module_init No platform data available.
[ 3.569832] brcmfmac 0002:21:00.0: enabling device (0000 -> 0002)
[ 3.569888] brcmfmac: brcmf_chip_recognition found AXI chip: BCM43752/2
[ 3.570843] brcmfmac: brcmf_chip_cores_check [1 ] core 0x800:65
base 0x18000000 wrap 0x18100000
[ 3.570848] brcmfmac: brcmf_chip_cores_check [2 ] core 0x812:82
base 0x18001000 wrap 0x18101000
[ 3.570853] brcmfmac: brcmf_chip_cores_check [3 ] core 0x83e:16
base 0x18002000 wrap 0x18102000
[ 3.570857] brcmfmac: brcmf_chip_cores_check [4 ] core 0x83c:65
base 0x18003000 wrap 0x18103000
[ 3.570860] brcmfmac: brcmf_chip_cores_check [5 ] core 0x829:26
base 0x18004000 wrap 0x18104000
[ 3.570864] brcmfmac: brcmf_chip_cores_check [6 ] core 0x844:2
base 0x18005000 wrap 0x18105000
[ 3.570868] brcmfmac: brcmf_chip_cores_check [7 ] core 0x840:15
base 0x18010000 wrap 0x00000000
[ 3.570872] brcmfmac: brcmf_chip_cores_check [8 ] core 0x827:35
base 0x18012000 wrap 0x00000000
[ 3.570876] brcmfmac: brcmf_chip_cores_check [9 ] core 0x240:0
base 0x00000000 wrap 0x1810a000
[ 3.570880] brcmfmac: brcmf_chip_cores_check [10] core 0x135:0
base 0x00000000 wrap 0x18112000
[ 3.570884] brcmfmac: brcmf_chip_cores_check [11] core 0x135:0
base 0x00000000 wrap 0x18113000
[ 3.570888] brcmfmac: brcmf_chip_set_passive Enter
[ 3.678925] brcmfmac: brcmf_chip_set_passive Enter
[ 3.679441] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.679443] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.679471] brcmfmac: brcmf_chip_setup ccrev=65, pmurev=35,
pmucaps=0x84565f23
[ 3.679477] brcmfmac: brcmf_get_module_param Enter, bus=2,
chip=43752, rev=2
[ 3.679495] brcmfmac: brcmf_alloc Enter
[ 3.679512] brcmfmac: brcmf_fw_alloc_request: using
brcm/fw_bcm43752a2_pcie for chip BCM43752/2
[ 3.679517] brcmfmac: brcmf_fw_get_firmwares enter: dev=0002:21:00.0
[ 3.679522] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin
[ 3.679826] brcmfmac 0002:21:00.0: Direct firmware load for
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.bin failed with error -2
[ 3.684423] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.bin found
[ 3.684433] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.txt
[ 3.684768] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.txt found
[ 3.684772] brcmfmac: brcmf_fw_request_nvram_done enter: dev=0002:21:00.0
[ 3.684947] brcmfmac: brcmf_fw_request_nvram_done nvram
000000007fae8cbc len 6040
[ 3.684955] brcmfmac: brcm_alt_fw_path FW alt path:
brcm/fw_bcm43752a2_pcie.rockchip,rk3588-evb1-v10.clm_blob
[ 3.685347] brcmfmac: brcmf_fw_complete_request firmware
brcm/fw_bcm43752a2_pcie.clm_blob found
[ 3.685625] brcmfmac: brcmf_chip_tcm_rambase: 43752 OK
[ 3.685627] brcmfmac: brcmf_chip_get_raminfo RAM: base=0x170000
size=1310720 (0x140000) sr=0 (0x0)
[ 3.779392] brcmfmac: Download NVRAM brcm/fw_bcm43752a2_pcie.txt
[ 3.780001] brcmfmac: brcmf_pcie_write_entropy_bytes: written 128
entropy bytes
[ 3.780016] brcmfmac: Bring ARM in running state
[ 3.780018] brcmfmac: brcmf_chip_set_active Enter
[ 3.780229] brcmfmac: Wait for FW init (0xfa1a05e5)
[ 3.995343] brcmfmac: brcmf_attach Enter
[ 3.995348] brcmfmac: brcmf_proto_attach Enter
[ 3.997068] brcmfmac: brcmf_fweh_register event handler registered
for PSM_WATCHDOG
[ 3.997073] brcmfmac: brcmf_bus_started
[ 3.997076] brcmfmac: brcmf_add_if Enter, bsscfgidx=0, ifidx=0
[ 3.997080] brcmfmac: brcmf_add_if allocate netdev interface
[ 3.997096] brcmfmac: brcmf_add_if ==== pid:bf, if:wlan%d
(00:00:00:00:00:00) created ===
[ 3.997100] brcmfmac: brcmf_bus_change_state 0 -> 1
[ 3.999755] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=cur_etheraddr, len=6, err=0
[ 4.000059] brcmfmac: brcmf_fil_cmd_data_get ifidx=0, cmd=98, len=68,
err=0
[ 4.000077] brcmfmac: brcmf_c_process_clm_blob Enter
[ 4.000083] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.000583] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.001002] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.001432] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.001852] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.002282] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.002713] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.003137] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.003569] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.003989] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.004420] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.004883] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.005163] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.005583] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.006008] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.006479] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.006864] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.007286] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.007718] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.008138] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.008568] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.008999] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=1415
[ 4.009417] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=clmload,
len=208
[ 4.012350] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=ver,
len=256, err=0
[ 4.012368] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM43752/2
wl0: Jul 12 2022 18:54:54 version 18.35.387.23.146 (g412cc5ec) FWID
01-93c53be6
[ 4.012791] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=clmver,
len=256, err=0
[ 4.012807] brcmfmac: brcmf_c_preinit_dcmds CLM version = API: 20.0
Data: 9.9.12_SS Compiler: 1.53.0 ClmImport: 1.52.0 Customization:
V15_190709 Creation: 2021-10-12 17:03:28
[ 4.012810] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=mpc, len=4
[ 4.013199] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=join_pref, len=8
[ 4.013759] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=event_msgs, len=18, err=0
[ 4.013770] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=event_msgs, len=18
[ 4.014108] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=185, value=40
[ 4.014506] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=187, value=40
[ 4.014930] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=txbf, len=4
[ 4.015732] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=cap,
len=768, err=0
[ 4.015749] brcmfmac: brcmf_feat_firmware_capabilities [ ap sta wme
802.11d 802.11h cqa cac dualband ampdu ampdu_tx ampdu_rx amsdurx amsdutx
tdls radio_pwrsave p2p proptxstatus mchan vht-prop-rates
single-user-beamformer multi-user-beamformee single-user-beamformee
dfrts txpwrcache stbc-tx stbc-rx-1ss mfp ndoe wnm bsstrans txcap epno
scanmac cptlv-4 ve fbtoverds fbt_adpt ecounters event_log cdef txcap ocl
hchk logtrace d11status scancache apf ifver idauth ifst fie tvpm
lpr_scan bkoff_evt etd_infoclm_restrict evt_ext btc_wifi_prot monitor
mbo estm d3cbuf sae extsae wrxf0ovfl ]
[ 4.015765] brcmfmac: brcmf_feat_firmware_capabilities enabling
feature: MCHAN
[ 4.015768] brcmfmac: brcmf_feat_firmware_capabilities enabling
feature: P2P
[ 4.015775] brcmfmac: brcmf_feat_firmware_capabilities enabling
feature: MONITOR
[ 4.015794] brcmfmac: brcmf_feat_firmware_capabilities enabling
feature: DOT11H
[ 4.015802] brcmfmac: brcmf_feat_firmware_capabilities enabling
feature: SAE
[ 4.015808] brcmfmac: brcmf_feat_firmware_capabilities enabling
feature: FWAUTH
[ 4.015810] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=pfn_gscan_cfg, len=18
[ 4.016360] brcmfmac: brcmf_fil_cmd_data Firmware error: BCME_VERSION
(-37)
[ 4.016364] brcmfmac: brcmf_feat_iovar_data_set enabling feature: GSCAN
[ 4.016731] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=pfn,
len=4, err=0
[ 4.016739] brcmfmac: brcmf_feat_iovar_int_get enabling feature: PNO
[ 4.017167] brcmfmac: brcmf_fil_cmd_data Firmware error:
BCME_UNSUPPORTED (-23)
[ 4.017170] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=wowl,
len=4, err=-23
[ 4.017177] brcmfmac: brcmf_feat_iovar_int_get WOWL feature check
failed: -23
[ 4.017456] brcmfmac: brcmf_fil_cmd_data Firmware error:
BCME_UNSUPPORTED (-23)
[ 4.017459] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=rsdb_mode, len=4, err=-23
[ 4.017467] brcmfmac: brcmf_feat_iovar_int_get RSDB feature check
failed: -23
[ 4.017823] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=tdls_enable, len=4, err=0
[ 4.017831] brcmfmac: brcmf_feat_iovar_int_get enabling feature: TDLS
[ 4.018220] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=mfp,
len=4, err=0
[ 4.018228] brcmfmac: brcmf_feat_iovar_int_get enabling feature: MFP
[ 4.018611] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=pfn_macaddr, len=8, err=0
[ 4.018933] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=sup_wpa,
len=4, err=0
[ 4.018942] brcmfmac: brcmf_feat_iovar_int_get enabling feature: FWSUP
[ 4.018947] brcmfmac: brcmf_alloc_vif allocating virtual interface
(size=4864)
[ 4.018961] brcmfmac: brcmf_fweh_register event handler registered
for LINK
[ 4.018964] brcmfmac: brcmf_fweh_register event handler registered
for DEAUTH_IND
[ 4.018966] brcmfmac: brcmf_fweh_register event handler registered
for DEAUTH
[ 4.018968] brcmfmac: brcmf_fweh_register event handler registered
for DISASSOC_IND
[ 4.018970] brcmfmac: brcmf_fweh_register event handler registered
for ASSOC_IND
[ 4.018972] brcmfmac: brcmf_fweh_register event handler registered
for REASSOC_IND
[ 4.018974] brcmfmac: brcmf_fweh_register event handler registered
for ROAM
[ 4.018976] brcmfmac: brcmf_fweh_register event handler registered
for MIC_ERROR
[ 4.018979] brcmfmac: brcmf_fweh_register event handler registered
for SET_SSID
[ 4.018981] brcmfmac: brcmf_fweh_register event handler registered
for PFN_NET_FOUND
[ 4.018983] brcmfmac: brcmf_fweh_register event handler registered for IF
[ 4.018986] brcmfmac: brcmf_fweh_register event handler registered
for P2P_PROBEREQ_MSG
[ 4.018988] brcmfmac: brcmf_fweh_register event handler registered
for P2P_DISC_LISTEN_COMPLETE
[ 4.018990] brcmfmac: brcmf_fweh_register event handler registered
for ACTION_FRAME_RX
[ 4.018993] brcmfmac: brcmf_fweh_register event handler registered
for ACTION_FRAME_COMPLETE
[ 4.018995] brcmfmac: brcmf_fweh_register event handler registered
for ACTION_FRAME_OFF_CHAN_COMPLETE
[ 4.018997] brcmfmac: brcmf_fweh_register event handler registered
for PSK_SUP
[ 4.019000] brcmfmac: brcmf_fweh_register event handler registered
for RSSI
[ 4.019002] brcmfmac: brcmf_fweh_register event handler registered
for ESCAN_RESULT
[ 4.019441] brcmfmac: brcmf_fil_cmd_int_get ifidx=0, cmd=1, value=2
[ 4.019748] brcmfmac: brcmf_fil_cmd_data_get ifidx=0, cmd=140,
len=12, err=0
[ 4.019763] brcmfmac: brcmf_cfg80211_attach Registering custom regulatory
[ 4.021256] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=vhtmode,
len=4, err=0
[ 4.021585] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=nmode,
len=4, err=0
[ 4.021915] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=bw_cap,
len=4, err=0
[ 4.022278] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=bw_cap,
len=4, err=0
[ 4.022286] brcmfmac: brcmf_setup_wiphybands nmode=1, vhtmode=1,
bw_cap=(1, 7)
[ 4.022530] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=rxchain,
len=4, err=0
[ 4.022538] brcmfmac: brcmf_setup_wiphybands nchain=2
[ 4.024274] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspecs, len=1536, err=0
[ 4.024793] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.025312] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.025832] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.026188] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.026718] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.027264] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.027782] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.028302] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.028822] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.029341] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.029698] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.030232] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.030589] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.031116] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.031660] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.032185] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.032575] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.032937] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.033462] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.033823] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.034358] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.034930] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.035455] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.036001] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.036368] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.036885] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.037242] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.037799] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.038161] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.038521] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.039049] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.039405] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.039931] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.040291] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.040816] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.041175] brcmfmac: brcmf_fil_bsscfg_data_get ifidx=0, bsscfgidx=0,
name=per_chan_info, len=4, err=0
[ 4.041578] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=txstreams, len=4, err=0
[ 4.041956] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=txbf_bfe_cap, len=4, err=0
[ 4.042320] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=txbf_bfr_cap, len=4, err=0
[ 4.042670] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=bw_cap,
len=4, err=0
[ 4.042678] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=bw_cap,
len=8
[ 4.043445] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspecs, len=1536, err=0
[ 4.043464] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=obss_coex, len=4
[ 4.043822] brcmfmac: brcmf_fweh_activate_events enable event SET_SSID
[ 4.043825] brcmfmac: brcmf_fweh_activate_events enable event DEAUTH
[ 4.043827] brcmfmac: brcmf_fweh_activate_events enable event DEAUTH_IND
[ 4.043829] brcmfmac: brcmf_fweh_activate_events enable event ASSOC_IND
[ 4.043831] brcmfmac: brcmf_fweh_activate_events enable event REASSOC_IND
[ 4.043833] brcmfmac: brcmf_fweh_activate_events enable event
DISASSOC_IND
[ 4.043836] brcmfmac: brcmf_fweh_activate_events enable event LINK
[ 4.043838] brcmfmac: brcmf_fweh_activate_events enable event MIC_ERROR
[ 4.043840] brcmfmac: brcmf_fweh_activate_events enable event ROAM
[ 4.043842] brcmfmac: brcmf_fweh_activate_events enable event
PFN_NET_FOUND
[ 4.043845] brcmfmac: brcmf_fweh_activate_events enable event
PSM_WATCHDOG
[ 4.043847] brcmfmac: brcmf_fweh_activate_events enable event PSK_SUP
[ 4.043849] brcmfmac: brcmf_fweh_activate_events enable event IF
[ 4.043851] brcmfmac: brcmf_fweh_activate_events enable event
P2P_DISC_LISTEN_COMPLETE
[ 4.043853] brcmfmac: brcmf_fweh_activate_events enable event RSSI
[ 4.043856] brcmfmac: brcmf_fweh_activate_events enable event
ACTION_FRAME_COMPLETE
[ 4.043858] brcmfmac: brcmf_fweh_activate_events enable event
ESCAN_RESULT
[ 4.043860] brcmfmac: brcmf_fweh_activate_events enable event
ACTION_FRAME_OFF_CHAN_COMPLETE
[ 4.043862] brcmfmac: brcmf_fweh_activate_events enable event
P2P_PROBEREQ_MSG
[ 4.043864] brcmfmac: brcmf_fweh_activate_events enable event
ACTION_FRAME_RX
[ 4.043867] brcmfmac: brcmf_fweh_activate_events enable event IF
[ 4.043869] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=event_msgs, len=18
[ 4.044164] brcmfmac: brcmf_btcoex_attach enter
[ 4.044167] brcmfmac: brcmf_pno_attach enter
[ 4.044171] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=tdls_enable, len=4
[ 4.044520] brcmfmac: brcmf_fweh_register event handler registered
for TDLS_PEER_EVENT
[ 4.044523] brcmfmac: brcmf_fweh_activate_events enable event SET_SSID
[ 4.044525] brcmfmac: brcmf_fweh_activate_events enable event DEAUTH
[ 4.044527] brcmfmac: brcmf_fweh_activate_events enable event DEAUTH_IND
[ 4.044529] brcmfmac: brcmf_fweh_activate_events enable event ASSOC_IND
[ 4.044531] brcmfmac: brcmf_fweh_activate_events enable event REASSOC_IND
[ 4.044533] brcmfmac: brcmf_fweh_activate_events enable event
DISASSOC_IND
[ 4.044536] brcmfmac: brcmf_fweh_activate_events enable event LINK
[ 4.044537] brcmfmac: brcmf_fweh_activate_events enable event MIC_ERROR
[ 4.044540] brcmfmac: brcmf_fweh_activate_events enable event ROAM
[ 4.044542] brcmfmac: brcmf_fweh_activate_events enable event
PFN_NET_FOUND
[ 4.044544] brcmfmac: brcmf_fweh_activate_events enable event
PSM_WATCHDOG
[ 4.044546] brcmfmac: brcmf_fweh_activate_events enable event PSK_SUP
[ 4.044548] brcmfmac: brcmf_fweh_activate_events enable event IF
[ 4.044550] brcmfmac: brcmf_fweh_activate_events enable event
P2P_DISC_LISTEN_COMPLETE
[ 4.044552] brcmfmac: brcmf_fweh_activate_events enable event RSSI
[ 4.044554] brcmfmac: brcmf_fweh_activate_events enable event
ACTION_FRAME_COMPLETE
[ 4.044557] brcmfmac: brcmf_fweh_activate_events enable event
ESCAN_RESULT
[ 4.044559] brcmfmac: brcmf_fweh_activate_events enable event
ACTION_FRAME_OFF_CHAN_COMPLETE
[ 4.044561] brcmfmac: brcmf_fweh_activate_events enable event
P2P_PROBEREQ_MSG
[ 4.044564] brcmfmac: brcmf_fweh_activate_events enable event
ACTION_FRAME_RX
[ 4.044566] brcmfmac: brcmf_fweh_activate_events enable event
TDLS_PEER_EVENT
[ 4.044568] brcmfmac: brcmf_fweh_activate_events enable event IF
[ 4.044570] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=event_msgs, len=18
[ 4.044864] brcmfmac: brcmf_net_attach Enter, bsscfgidx=0
mac=d4:9c:dd:f5:bf:7e
[ 4.046336] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspec, len=4, err=0
[ 4.046350] brcmfmac: brcmf_cfg80211_get_tx_power Enter
[ 4.046352] brcmfmac: check_vif_up device is not ready : status (0)
[ 4.046386] brcmfmac: brcmf_net_attach wlan0: Broadcom Dongle Host Driver
[ 4.052221] brcmfmac 0002:21:00.0 wlP2p33s0: renamed from wlan0
[ 4.629677] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspec, len=4, err=0
[ 4.629688] brcmfmac: brcmf_cfg80211_get_tx_power Enter
[ 4.629689] brcmfmac: check_vif_up device is not ready : status (0)
[ 4.629733] brcmfmac: brcmf_cfg80211_dump_station Enter, idx 0
[ 4.630174] brcmfmac: brcmf_fil_cmd_data_get ifidx=0, cmd=159,
len=772, err=0
[ 4.970573] brcmfmac: brcmf_netdev_open Enter, bsscfgidx=0
[ 4.972550] brcmfmac: brcmf_fil_cmd_data Firmware error:
BCME_UNSUPPORTED (-23)
[ 4.972552] brcmfmac: brcmf_fil_iovar_data_get ifidx=0, name=toe_ol,
len=4, err=-52
[ 4.972558] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=2, value=0
[ 4.999772] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=185, value=40
[ 4.999812] brcmfmac: brcmf_fweh_event_worker event IF (54) ifidx 0
bsscfg 0 addr d4:9c:dd:f5:bf:7e
[ 4.999819] brcmfmac: brcmf_fweh_event_worker version 2 flags 0
status 0 reason 0
[ 4.999829] brcmfmac: brcmf_fweh_handle_if_event action: 1 ifidx: 0
bsscfgidx: 0 flags: 0 role: 0
[ 4.999832] brcmfmac: brcmf_fweh_handle_if_event adding wl0
(d4:9c:dd:f5:bf:7e)
[ 4.999835] brcmfmac: brcmf_add_if Enter, bsscfgidx=0, ifidx=0
[ 4.999838] brcmfmac: brcmf_add_if netdev:wlP2p33s0 ignore IF event
[ 5.000315] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=187, value=40
[ 5.000563] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=258, value=120
[ 5.000922] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=86, value=2
[ 5.001170] brcmfmac: brcmf_config_dongle power save set to enabled
[ 5.001172] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=bcn_timeout, len=4
[ 5.001399] brcmfmac: brcmf_dongle_roam Internal Roaming = On
[ 5.001401] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=roam_off, len=4
[ 5.001737] brcmfmac: brcmf_fil_cmd_data_set ifidx=0, cmd=55, len=8
[ 5.001866] brcmfmac: brcmf_fil_cmd_data_set ifidx=0, cmd=57, len=8
[ 5.002110] brcmfmac: brcmf_cfg80211_change_iface Enter, bsscfgidx=0,
type=2
[ 5.002112] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=20, value=1
[ 5.002512] brcmfmac: brcmf_cfg80211_change_iface IF Type = Infra
[ 5.002514] brcmfmac: brcmf_cfg80211_change_iface Exit
[ 5.002515] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=arp_ol,
len=4
[ 5.002722] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=arpoe, len=4
[ 5.003040] brcmfmac: brcmf_configure_arp_nd_offload successfully
configured (1) ARP offload to 0x9
[ 5.003042] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=ndoe, len=4
[ 5.003358] brcmfmac: brcmf_configure_arp_nd_offload successfully
configured (1) ND offload to 0x9
[ 5.003360] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=219, value=1
[ 5.003606] brcmfmac: _brcmf_set_multicast_list Enter, bsscfgidx=0
[ 5.003608] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=mcast_list, len=10
[ 5.003659] brcmfmac: brcmf_cfg80211_set_power_mgmt Enter
[ 5.003660] brcmfmac: brcmf_cfg80211_set_power_mgmt power save enabled
[ 5.003992] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=86, value=2
[ 5.004224] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=allmulti, len=4
[ 5.004614] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=pm2_sleep_ret, len=4
[ 5.004824] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=10, value=0
[ 5.004827] brcmfmac: brcmf_cfg80211_set_power_mgmt Exit
[ 5.004954] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=arp_ol,
len=4
[ 5.005270] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=arpoe, len=4
[ 5.005474] brcmfmac: brcmf_configure_arp_nd_offload successfully
configured (1) ARP offload to 0x9
[ 5.005475] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=ndoe, len=4
[ 5.005788] brcmfmac: brcmf_configure_arp_nd_offload successfully
configured (1) ND offload to 0x9
[ 5.005790] brcmfmac: _brcmf_set_multicast_list Enter, bsscfgidx=0
[ 5.005792] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=mcast_list, len=16
[ 5.006050] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=allmulti, len=4
[ 5.006293] brcmfmac: brcmf_fil_cmd_int_set ifidx=0, cmd=10, value=0
[ 5.006420] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=arp_ol,
len=4
[ 5.006626] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=arpoe, len=4
[ 5.006835] brcmfmac: brcmf_configure_arp_nd_offload successfully
configured (1) ARP offload to 0x9
[ 5.006837] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=ndoe, len=4
[ 5.007148] brcmfmac: brcmf_configure_arp_nd_offload successfully
configured (1) ND offload to 0x9
[ 5.008354] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspec, len=4, err=0
[ 5.008362] brcmfmac: brcmf_cfg80211_get_tx_power Enter
[ 5.008607] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=qtxpower, len=4, err=0
[ 5.008610] brcmfmac: brcmf_cfg80211_get_tx_power Exit (0x7f 31)
[ 5.008859] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspec, len=4, err=0
[ 5.008862] brcmfmac: brcmf_cfg80211_get_tx_power Enter
[ 5.009112] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=qtxpower, len=4, err=0
[ 5.009114] brcmfmac: brcmf_cfg80211_get_tx_power Exit (0x7f 31)
[ 5.009404] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspec, len=4, err=0
[ 5.009408] brcmfmac: brcmf_cfg80211_get_tx_power Enter
[ 5.009693] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=qtxpower, len=4, err=0
[ 5.009695] brcmfmac: brcmf_cfg80211_get_tx_power Exit (0x7f 31)
[ 5.009851] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspec, len=4, err=0
[ 5.009855] brcmfmac: brcmf_cfg80211_get_tx_power Enter
[ 5.010102] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=qtxpower, len=4, err=0
[ 5.010105] brcmfmac: brcmf_cfg80211_get_tx_power Exit (0x7f 31)
[ 5.010732] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=chanspec, len=4, err=0
[ 5.010737] brcmfmac: brcmf_cfg80211_get_tx_power Enter
[ 5.010984] brcmfmac: brcmf_fil_iovar_data_get ifidx=0,
name=qtxpower, len=4, err=0
[ 5.010987] brcmfmac: brcmf_cfg80211_get_tx_power Exit (0x7f 31)
[ 5.035480] brcmfmac: brcmf_cfg80211_flush_pmksa Enter
[ 5.035489] brcmfmac: brcmf_fil_iovar_data_set ifidx=0,
name=pmkid_info, len=356
[ 5.035999] brcmfmac: brcmf_cfg80211_flush_pmksa Exit
[ 5.083266] brcmfmac: brcmf_cfg80211_scan Enter
[ 5.083271] brcmfmac: brcmf_cfg80211_scan START ESCAN
[ 5.083272] brcmfmac: brcmf_vif_set_mgmt_ie bsscfgidx 0, pktflag : 0x10
[ 5.083275] brcmfmac: brcmf_do_escan Enter
[ 5.083277] brcmfmac: brcmf_run_escan E-SCAN START
[ 5.083278] brcmfmac: brcmf_escan_prep ### List of channelspecs to
scan ### 36
[ 5.083280] brcmfmac: brcmf_escan_prep Chan : 1, Channel spec: 1001
[ 5.083281] brcmfmac: brcmf_escan_prep Chan : 2, Channel spec: 1002
[ 5.083282] brcmfmac: brcmf_escan_prep Chan : 3, Channel spec: 1003
[ 5.083283] brcmfmac: brcmf_escan_prep Chan : 4, Channel spec: 1004
[ 5.083284] brcmfmac: brcmf_escan_prep Chan : 5, Channel spec: 1005
[ 5.083285] brcmfmac: brcmf_escan_prep Chan : 6, Channel spec: 1006
[ 5.083286] brcmfmac: brcmf_escan_prep Chan : 7, Channel spec: 1007
[ 5.083287] brcmfmac: brcmf_escan_prep Chan : 8, Channel spec: 1008
[ 5.083287] brcmfmac: brcmf_escan_prep Chan : 9, Channel spec: 1009
[ 5.083288] brcmfmac: brcmf_escan_prep Chan : 10, Channel spec: 100a
[ 5.083289] brcmfmac: brcmf_escan_prep Chan : 11, Channel spec: 100b
[ 5.083290] brcmfmac: brcmf_escan_prep Chan : 36, Channel spec: d024
[ 5.083291] brcmfmac: brcmf_escan_prep Chan : 40, Channel spec: d028
[ 5.083292] brcmfmac: brcmf_escan_prep Chan : 44, Channel spec: d02c
[ 5.083293] brcmfmac: brcmf_escan_prep Chan : 48, Channel spec: d030
[ 5.083294] brcmfmac: brcmf_escan_prep Chan : 52, Channel spec: d034
[ 5.083294] brcmfmac: brcmf_escan_prep Chan : 56, Channel spec: d038
[ 5.083295] brcmfmac: brcmf_escan_prep Chan : 60, Channel spec: d03c
[ 5.083296] brcmfmac: brcmf_escan_prep Chan : 64, Channel spec: d040
[ 5.083297] brcmfmac: brcmf_escan_prep Chan : 100, Channel spec: d064
[ 5.083298] brcmfmac: brcmf_escan_prep Chan : 104, Channel spec: d068
[ 5.083299] brcmfmac: brcmf_escan_prep Chan : 108, Channel spec: d06c
[ 5.083300] brcmfmac: brcmf_escan_prep Chan : 112, Channel spec: d070
[ 5.083301] brcmfmac: brcmf_escan_prep Chan : 116, Channel spec: d074
[ 5.083301] brcmfmac: brcmf_escan_prep Chan : 120, Channel spec: d078
[ 5.083303] brcmfmac: brcmf_escan_prep Chan : 124, Channel spec: d07c
[ 5.083303] brcmfmac: brcmf_escan_prep Chan : 128, Channel spec: d080
[ 5.083304] brcmfmac: brcmf_escan_prep Chan : 132, Channel spec: d084
[ 5.083305] brcmfmac: brcmf_escan_prep Chan : 136, Channel spec: d088
[ 5.083306] brcmfmac: brcmf_escan_prep Chan : 140, Channel spec: d08c
[ 5.083307] brcmfmac: brcmf_escan_prep Chan : 144, Channel spec: d090
[ 5.083308] brcmfmac: brcmf_escan_prep Chan : 149, Channel spec: d095
[ 5.083309] brcmfmac: brcmf_escan_prep Chan : 153, Channel spec: d099
[ 5.083310] brcmfmac: brcmf_escan_prep Chan : 157, Channel spec: d09d
[ 5.083310] brcmfmac: brcmf_escan_prep Chan : 161, Channel spec: d0a1
[ 5.083311] brcmfmac: brcmf_escan_prep Chan : 165, Channel spec: d0a5
[ 5.083312] brcmfmac: brcmf_escan_prep ### List of SSIDs to scan ### 1
[ 5.083313] brcmfmac: brcmf_escan_prep 0: Broadcast scan
[ 5.083314] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=escan,
len=180
[ 5.083722] brcmfmac: brcmf_fil_cmd_data Firmware error: BCME_VERSION
(-37)
[ 8.316080] brcmfmac: brcmf_cfg80211_scan Enter
[ 8.316085] brcmfmac: brcmf_cfg80211_scan START ESCAN
[ 8.316087] brcmfmac: brcmf_vif_set_mgmt_ie bsscfgidx 0, pktflag : 0x10
[ 8.316091] brcmfmac: brcmf_do_escan Enter
[ 8.316092] brcmfmac: brcmf_run_escan E-SCAN START
[ 8.316093] brcmfmac: brcmf_escan_prep ### List of channelspecs to
scan ### 36
[ 8.316095] brcmfmac: brcmf_escan_prep Chan : 1, Channel spec: 1001
[ 8.316096] brcmfmac: brcmf_escan_prep Chan : 2, Channel spec: 1002
[ 8.316097] brcmfmac: brcmf_escan_prep Chan : 3, Channel spec: 1003
[ 8.316098] brcmfmac: brcmf_escan_prep Chan : 4, Channel spec: 1004
[ 8.316099] brcmfmac: brcmf_escan_prep Chan : 5, Channel spec: 1005
[ 8.316100] brcmfmac: brcmf_escan_prep Chan : 6, Channel spec: 1006
[ 8.316101] brcmfmac: brcmf_escan_prep Chan : 7, Channel spec: 1007
[ 8.316102] brcmfmac: brcmf_escan_prep Chan : 8, Channel spec: 1008
[ 8.316103] brcmfmac: brcmf_escan_prep Chan : 9, Channel spec: 1009
[ 8.316104] brcmfmac: brcmf_escan_prep Chan : 10, Channel spec: 100a
[ 8.316104] brcmfmac: brcmf_escan_prep Chan : 11, Channel spec: 100b
[ 8.316105] brcmfmac: brcmf_escan_prep Chan : 36, Channel spec: d024
[ 8.316106] brcmfmac: brcmf_escan_prep Chan : 40, Channel spec: d028
[ 8.316107] brcmfmac: brcmf_escan_prep Chan : 44, Channel spec: d02c
[ 8.316108] brcmfmac: brcmf_escan_prep Chan : 48, Channel spec: d030
[ 8.316109] brcmfmac: brcmf_escan_prep Chan : 52, Channel spec: d034
[ 8.316110] brcmfmac: brcmf_escan_prep Chan : 56, Channel spec: d038
[ 8.316111] brcmfmac: brcmf_escan_prep Chan : 60, Channel spec: d03c
[ 8.316112] brcmfmac: brcmf_escan_prep Chan : 64, Channel spec: d040
[ 8.316113] brcmfmac: brcmf_escan_prep Chan : 100, Channel spec: d064
[ 8.316114] brcmfmac: brcmf_escan_prep Chan : 104, Channel spec: d068
[ 8.316115] brcmfmac: brcmf_escan_prep Chan : 108, Channel spec: d06c
[ 8.316115] brcmfmac: brcmf_escan_prep Chan : 112, Channel spec: d070
[ 8.316116] brcmfmac: brcmf_escan_prep Chan : 116, Channel spec: d074
[ 8.316117] brcmfmac: brcmf_escan_prep Chan : 120, Channel spec: d078
[ 8.316118] brcmfmac: brcmf_escan_prep Chan : 124, Channel spec: d07c
[ 8.316119] brcmfmac: brcmf_escan_prep Chan : 128, Channel spec: d080
[ 8.316120] brcmfmac: brcmf_escan_prep Chan : 132, Channel spec: d084
[ 8.316121] brcmfmac: brcmf_escan_prep Chan : 136, Channel spec: d088
[ 8.316122] brcmfmac: brcmf_escan_prep Chan : 140, Channel spec: d08c
[ 8.316123] brcmfmac: brcmf_escan_prep Chan : 144, Channel spec: d090
[ 8.316124] brcmfmac: brcmf_escan_prep Chan : 149, Channel spec: d095
[ 8.316125] brcmfmac: brcmf_escan_prep Chan : 153, Channel spec: d099
[ 8.316125] brcmfmac: brcmf_escan_prep Chan : 157, Channel spec: d09d
[ 8.316126] brcmfmac: brcmf_escan_prep Chan : 161, Channel spec: d0a1
[ 8.316127] brcmfmac: brcmf_escan_prep Chan : 165, Channel spec: d0a5
[ 8.316128] brcmfmac: brcmf_escan_prep ### List of SSIDs to scan ### 1
[ 8.316129] brcmfmac: brcmf_escan_prep 0: Broadcast scan
[ 8.316130] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=escan,
len=180
[ 8.317949] brcmfmac: brcmf_fil_cmd_data Firmware error: BCME_VERSION
(-37)
[ 17.202228] brcmfmac: brcmf_cfg80211_scan Enter
[ 17.202237] brcmfmac: brcmf_cfg80211_scan START ESCAN
[ 17.202240] brcmfmac: brcmf_vif_set_mgmt_ie bsscfgidx 0, pktflag : 0x10
[ 17.202247] brcmfmac: brcmf_do_escan Enter
[ 17.202250] brcmfmac: brcmf_run_escan E-SCAN START
[ 17.202253] brcmfmac: brcmf_escan_prep ### List of channelspecs to
scan ### 36
[ 17.202257] brcmfmac: brcmf_escan_prep Chan : 1, Channel spec: 1001
[ 17.202260] brcmfmac: brcmf_escan_prep Chan : 2, Channel spec: 1002
[ 17.202263] brcmfmac: brcmf_escan_prep Chan : 3, Channel spec: 1003
[ 17.202265] brcmfmac: brcmf_escan_prep Chan : 4, Channel spec: 1004
[ 17.202267] brcmfmac: brcmf_escan_prep Chan : 5, Channel spec: 1005
[ 17.202270] brcmfmac: brcmf_escan_prep Chan : 6, Channel spec: 1006
[ 17.202272] brcmfmac: brcmf_escan_prep Chan : 7, Channel spec: 1007
[ 17.202275] brcmfmac: brcmf_escan_prep Chan : 8, Channel spec: 1008
[ 17.202277] brcmfmac: brcmf_escan_prep Chan : 9, Channel spec: 1009
[ 17.202280] brcmfmac: brcmf_escan_prep Chan : 10, Channel spec: 100a
[ 17.202283] brcmfmac: brcmf_escan_prep Chan : 11, Channel spec: 100b
[ 17.202285] brcmfmac: brcmf_escan_prep Chan : 36, Channel spec: d024
[ 17.202288] brcmfmac: brcmf_escan_prep Chan : 40, Channel spec: d028
[ 17.202290] brcmfmac: brcmf_escan_prep Chan : 44, Channel spec: d02c
[ 17.202293] brcmfmac: brcmf_escan_prep Chan : 48, Channel spec: d030
[ 17.202295] brcmfmac: brcmf_escan_prep Chan : 52, Channel spec: d034
[ 17.202298] brcmfmac: brcmf_escan_prep Chan : 56, Channel spec: d038
[ 17.202300] brcmfmac: brcmf_escan_prep Chan : 60, Channel spec: d03c
[ 17.202303] brcmfmac: brcmf_escan_prep Chan : 64, Channel spec: d040
[ 17.202305] brcmfmac: brcmf_escan_prep Chan : 100, Channel spec: d064
[ 17.202308] brcmfmac: brcmf_escan_prep Chan : 104, Channel spec: d068
[ 17.202311] brcmfmac: brcmf_escan_prep Chan : 108, Channel spec: d06c
[ 17.202313] brcmfmac: brcmf_escan_prep Chan : 112, Channel spec: d070
[ 17.202316] brcmfmac: brcmf_escan_prep Chan : 116, Channel spec: d074
[ 17.202318] brcmfmac: brcmf_escan_prep Chan : 120, Channel spec: d078
[ 17.202321] brcmfmac: brcmf_escan_prep Chan : 124, Channel spec: d07c
[ 17.202323] brcmfmac: brcmf_escan_prep Chan : 128, Channel spec: d080
[ 17.202326] brcmfmac: brcmf_escan_prep Chan : 132, Channel spec: d084
[ 17.202328] brcmfmac: brcmf_escan_prep Chan : 136, Channel spec: d088
[ 17.202330] brcmfmac: brcmf_escan_prep Chan : 140, Channel spec: d08c
[ 17.202333] brcmfmac: brcmf_escan_prep Chan : 144, Channel spec: d090
[ 17.202335] brcmfmac: brcmf_escan_prep Chan : 149, Channel spec: d095
[ 17.202338] brcmfmac: brcmf_escan_prep Chan : 153, Channel spec: d099
[ 17.202340] brcmfmac: brcmf_escan_prep Chan : 157, Channel spec: d09d
[ 17.202343] brcmfmac: brcmf_escan_prep Chan : 161, Channel spec: d0a1
[ 17.202345] brcmfmac: brcmf_escan_prep Chan : 165, Channel spec: d0a5
[ 17.202348] brcmfmac: brcmf_escan_prep ### List of SSIDs to scan ### 1
[ 17.202350] brcmfmac: brcmf_escan_prep 0: Broadcast scan
[ 17.202352] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=escan,
len=180
[ 17.204223] brcmfmac: brcmf_fil_cmd_data Firmware error: BCME_VERSION
(-37)
[email protected]:~# [ 44.202645] ieee80211 phy0: brcmf_run_escan:
error (-52)
[ 44.202649] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)

-Andy


> Regards,
> Arend

2022-11-24 13:50:47

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/24/2022 1:54 PM, Andy Green wrote:
> [   17.202348] brcmfmac: brcmf_escan_prep ### List of SSIDs to scan ### 1
> [   17.202350] brcmfmac: brcmf_escan_prep 0: Broadcast scan
> [   17.202352] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=escan,
> len=180
> [   17.204223] brcmfmac: brcmf_fil_cmd_data Firmware error: BCME_VERSION
> (-37)
> [email protected]:~# [   44.202645] ieee80211 phy0: brcmf_run_escan:
> error (-52)
> [   44.202649] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)

The root cause is easy. Basically the brcmfmac driver is not on par with
the firmware. brcmfmac uses escan api version 1 and firmware expects
version 2. Fixing it is going to be more difficult. Things start to
diverge on firmware API level. I already had a patch on the shelf
supporting version 2, but it does not match the API your firmware expects.

I was working on supporting multiple vendors, but I need some rework.
The scanning issue confirms the need for per-vendor support. Hopefully I
can spend some time on it.

Regards,
Arend


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-11-24 14:24:03

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/24/22 13:47, Arend van Spriel wrote:
> On 11/24/2022 1:54 PM, Andy Green wrote:
>> [   17.202348] brcmfmac: brcmf_escan_prep ### List of SSIDs to scan ### 1
>> [   17.202350] brcmfmac: brcmf_escan_prep 0: Broadcast scan
>> [   17.202352] brcmfmac: brcmf_fil_iovar_data_set ifidx=0, name=escan,
>> len=180
>> [   17.204223] brcmfmac: brcmf_fil_cmd_data Firmware error:
>> BCME_VERSION (-37)
>> [email protected]:~# [   44.202645] ieee80211 phy0:
>> brcmf_run_escan: error (-52)
>> [   44.202649] ieee80211 phy0: brcmf_cfg80211_scan: scan error (-52)
>
> The root cause is easy. Basically the brcmfmac driver is not on par with
> the firmware. brcmfmac uses escan api version 1 and firmware expects
> version 2. Fixing it is going to be more difficult. Things start to

It seems so.

> diverge on firmware API level. I already had a patch on the shelf
> supporting version 2, but it does not match the API your firmware expects.
>
> I was working on supporting multiple vendors, but I need some rework.
> The scanning issue confirms the need for per-vendor support. Hopefully I
> can spend some time on it.

If it makes sense for you, I'm willing take a look at your alpha patch
with the idea to align with what the OOT driver does for scan, but I'm
not sure how intricate wading in there is going to get.

I'd need at least some guidance about how to detect which api stack it
wants dynamically if there's more than one "version 2" and likely some
hand-holding about firmware or soc-specific things in places.

-Andy

2022-11-24 14:40:34

by Andy Green

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf



On 11/24/22 14:16, Andy Green wrote:

>> diverge on firmware API level. I already had a patch on the shelf
>> supporting version 2, but it does not match the API your firmware
>> expects.
>>
>> I was working on supporting multiple vendors, but I need some rework.
>> The scanning issue confirms the need for per-vendor support. Hopefully
>> I can spend some time on it.
>
> If it makes sense for you, I'm willing take a look at your alpha patch
> with the idea to align with what the OOT driver does for scan, but I'm
> not sure how intricate wading in there is going to get.
>
> I'd need at least some guidance about how to detect which api stack it
> wants dynamically if there's more than one "version 2" and likely some
> hand-holding about firmware or soc-specific things in places.

... there's the additional quirk that my rebase on mainline of the OOT
driver starts up like mainline now, but doesn't return any scan results
as it is.

So aligning to that may only take us so far without some possibly
non-open knowledge about what the firmware really wants from its
perspective directly without referencing dhd.

-Andy

2022-11-29 19:59:38

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 11/17/2022 7:36 PM, Angus Ainslie wrote:
> On 2022-11-17 10:12, Arend Van Spriel wrote:
>> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]> wrote:
>>
>>> On 2022-11-17 04:01, Andy Green wrote:
>>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>>> Got a new version of downstream driver and firmware here:
>>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>>
>>>>> Thanks for that. The firmware indeed seems to originate from Synaptics
>>>>> and this module can be found on wifi-alliance product finder:
>>>>>
>>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>>
>>>>> We have no known contacts there and they do not seem to have much
>>>>> interest in open-source. In that gitlab repo I see they are providing
>>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>>
>>>>> Will see what I can do. First have to think on how to approach this
>>>>> failure to get most out of a debug effort. I do not have the module
>>>>> over here so we would need to go back and forth. The fact that the
>>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>>
>>>> Yes it seems that at some point with some firmware, SDIO has worked
>>>> for, eg, Angus.  And, it should work well with bchdhd (but that does
>>>> not seem to work for scan for me on v6.1-rc5 uplevel with the firmware
>>>> load changed to use kernel apis; but the firmware has no timeout on
>>>> bcmdhd). So it doesn't feel a million miles away.
>>>
>>> This is the firmware we have been using for SDIO. It was sent to us
>>> directly by SparkLAN so not sure how it compares to other variants.
>>>
>>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>>
>> It is a tad bit older:
>>
>> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
>> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
>> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
>> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>>
>
> Currently it's broken for hotspot mode but I haven't had the chance to
> see if that is FW or something further up the stack.

Hi Angus,

When you added support for this device in brcmfmac you marked it as a
Cypress device. Are you sure about that? AFAICT this device is coming
from Synaptics.

Regards,
Arend

> Angus
>
>> Thanks,
>> Arend
>>
>>>
>>> Angus
>>>
>>>> I have the hw, and (paid) time to help with debugging this "pair of
>>>> hands" style. I'm also a little familiar with the kernel code and can
>>>> try jiggle things autonomously if you can prompt what is interesting.
>>>>
>>>> -Andy
>>>>
>>>>> Regards,
>>>>> Arend


Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-12-03 13:59:15

by Angus Ainslie

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 2022-11-29 11:48, Arend van Spriel wrote:
> On 11/17/2022 7:36 PM, Angus Ainslie wrote:
>> On 2022-11-17 10:12, Arend Van Spriel wrote:
>>> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]> wrote:
>>>
>>>> On 2022-11-17 04:01, Andy Green wrote:
>>>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>>>> Got a new version of downstream driver and firmware here:
>>>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>>>
>>>>>> Thanks for that. The firmware indeed seems to originate from
>>>>>> Synaptics
>>>>>> and this module can be found on wifi-alliance product finder:
>>>>>>
>>>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>>>
>>>>>> We have no known contacts there and they do not seem to have much
>>>>>> interest in open-source. In that gitlab repo I see they are
>>>>>> providing
>>>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>>>
>>>>>> Will see what I can do. First have to think on how to approach
>>>>>> this
>>>>>> failure to get most out of a debug effort. I do not have the
>>>>>> module
>>>>>> over here so we would need to go back and forth. The fact that the
>>>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>>>
>>>>> Yes it seems that at some point with some firmware, SDIO has worked
>>>>> for, eg, Angus.  And, it should work well with bchdhd (but that
>>>>> does
>>>>> not seem to work for scan for me on v6.1-rc5 uplevel with the
>>>>> firmware
>>>>> load changed to use kernel apis; but the firmware has no timeout on
>>>>> bcmdhd). So it doesn't feel a million miles away.
>>>>
>>>> This is the firmware we have been using for SDIO. It was sent to us
>>>> directly by SparkLAN so not sure how it compares to other variants.
>>>>
>>>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>>>
>>> It is a tad bit older:
>>>
>>> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
>>> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
>>> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
>>> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>>>
>>
>> Currently it's broken for hotspot mode but I haven't had the chance to
>> see if that is FW or something further up the stack.
>
> Hi Angus,
>
> When you added support for this device in brcmfmac you marked it as a
> Cypress device. Are you sure about that? AFAICT this device is coming
> from Synaptics.
>

Hi Arend,

I assumed Cypress based on the fact that it most closely resembled the
4373 from the driver side.

How would I figure out whether it's Cypress or Synaptics ?

When I search the web for it I only find SparkLAN or "AMPAK technology"
references.

Thanks
Angus


> Regards,
> Arend
>
>> Angus
>>
>>> Thanks,
>>> Arend
>>>
>>>>
>>>> Angus
>>>>
>>>>> I have the hw, and (paid) time to help with debugging this "pair of
>>>>> hands" style. I'm also a little familiar with the kernel code and
>>>>> can
>>>>> try jiggle things autonomously if you can prompt what is
>>>>> interesting.
>>>>>
>>>>> -Andy
>>>>>
>>>>>> Regards,
>>>>>> Arend

2022-12-03 15:30:42

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On December 3, 2022 2:56:39 PM Angus Ainslie <[email protected]> wrote:

> On 2022-11-29 11:48, Arend van Spriel wrote:
>> On 11/17/2022 7:36 PM, Angus Ainslie wrote:
>>> On 2022-11-17 10:12, Arend Van Spriel wrote:
>>>> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]> wrote:
>>>>
>>>>> On 2022-11-17 04:01, Andy Green wrote:
>>>>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>>>>> Got a new version of downstream driver and firmware here:
>>>>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>>>>
>>>>>>> Thanks for that. The firmware indeed seems to originate from
>>>>>>> Synaptics
>>>>>>> and this module can be found on wifi-alliance product finder:
>>>>>>>
>>>>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>>>>
>>>>>>> We have no known contacts there and they do not seem to have much
>>>>>>> interest in open-source. In that gitlab repo I see they are
>>>>>>> providing
>>>>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>>>>
>>>>>>> Will see what I can do. First have to think on how to approach
>>>>>>> this
>>>>>>> failure to get most out of a debug effort. I do not have the
>>>>>>> module
>>>>>>> over here so we would need to go back and forth. The fact that the
>>>>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>>>>
>>>>>> Yes it seems that at some point with some firmware, SDIO has worked
>>>>>> for, eg, Angus. And, it should work well with bchdhd (but that
>>>>>> does
>>>>>> not seem to work for scan for me on v6.1-rc5 uplevel with the
>>>>>> firmware
>>>>>> load changed to use kernel apis; but the firmware has no timeout on
>>>>>> bcmdhd). So it doesn't feel a million miles away.
>>>>>
>>>>> This is the firmware we have been using for SDIO. It was sent to us
>>>>> directly by SparkLAN so not sure how it compares to other variants.
>>>>>
>>>>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>>>>
>>>> It is a tad bit older:
>>>>
>>>> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
>>>> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
>>>> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
>>>> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>>>
>>> Currently it's broken for hotspot mode but I haven't had the chance to
>>> see if that is FW or something further up the stack.
>>
>> Hi Angus,
>>
>> When you added support for this device in brcmfmac you marked it as a
>> Cypress device. Are you sure about that? AFAICT this device is coming
>> from Synaptics.
>
> Hi Arend,
>
> I assumed Cypress based on the fact that it most closely resembled the
> 4373 from the driver side.

Right. Both are based on the same chip architecture that they brought from
Broadcom so the similarities are not surprising.

> How would I figure out whether it's Cypress or Synaptics ?
>
> When I search the web for it I only find SparkLAN or "AMPAK technology"
> references.

Earlier my colleague Franky found this link to WFA certificate:

https://api.cert.wi-fi.org/api/certificate/download/public?variantId=103641

WFA and FCC are usually good sources of info. Also the product pages of the
vendors are worth checking. I have them bookmarked out of professional
interest ;-)

Regards,
Arend



Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-12-03 15:42:34

by Angus Ainslie

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 2022-12-03 07:22, Arend Van Spriel wrote:
> On December 3, 2022 2:56:39 PM Angus Ainslie <[email protected]> wrote:
>
>> On 2022-11-29 11:48, Arend van Spriel wrote:
>>> On 11/17/2022 7:36 PM, Angus Ainslie wrote:
>>>> On 2022-11-17 10:12, Arend Van Spriel wrote:
>>>>> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]>
>>>>> wrote:
>>>>>
>>>>>> On 2022-11-17 04:01, Andy Green wrote:
>>>>>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>>>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>>>>>> Got a new version of downstream driver and firmware here:
>>>>>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>>>>>
>>>>>>>> Thanks for that. The firmware indeed seems to originate from
>>>>>>>> Synaptics
>>>>>>>> and this module can be found on wifi-alliance product finder:
>>>>>>>>
>>>>>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>>>>>
>>>>>>>> We have no known contacts there and they do not seem to have
>>>>>>>> much
>>>>>>>> interest in open-source. In that gitlab repo I see they are
>>>>>>>> providing
>>>>>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>>>>>
>>>>>>>> Will see what I can do. First have to think on how to approach
>>>>>>>> this
>>>>>>>> failure to get most out of a debug effort. I do not have the
>>>>>>>> module
>>>>>>>> over here so we would need to go back and forth. The fact that
>>>>>>>> the
>>>>>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>>>>>
>>>>>>> Yes it seems that at some point with some firmware, SDIO has
>>>>>>> worked
>>>>>>> for, eg, Angus. And, it should work well with bchdhd (but that
>>>>>>> does
>>>>>>> not seem to work for scan for me on v6.1-rc5 uplevel with the
>>>>>>> firmware
>>>>>>> load changed to use kernel apis; but the firmware has no timeout
>>>>>>> on
>>>>>>> bcmdhd). So it doesn't feel a million miles away.
>>>>>>
>>>>>> This is the firmware we have been using for SDIO. It was sent to
>>>>>> us
>>>>>> directly by SparkLAN so not sure how it compares to other
>>>>>> variants.
>>>>>>
>>>>>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>>>>>
>>>>> It is a tad bit older:
>>>>>
>>>>> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
>>>>> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
>>>>> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
>>>>> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>>>>
>>>> Currently it's broken for hotspot mode but I haven't had the chance
>>>> to
>>>> see if that is FW or something further up the stack.
>>>
>>> Hi Angus,
>>>
>>> When you added support for this device in brcmfmac you marked it as a
>>> Cypress device. Are you sure about that? AFAICT this device is coming
>>> from Synaptics.
>>
>> Hi Arend,
>>
>> I assumed Cypress based on the fact that it most closely resembled the
>> 4373 from the driver side.
>
> Right. Both are based on the same chip architecture that they brought
> from Broadcom so the similarities are not surprising.
>
>> How would I figure out whether it's Cypress or Synaptics ?
>>
>> When I search the web for it I only find SparkLAN or "AMPAK
>> technology"
>> references.
>
> Earlier my colleague Franky found this link to WFA certificate:
>
> https://api.cert.wi-fi.org/api/certificate/download/public?variantId=103641
>
> WFA and FCC are usually good sources of info. Also the product pages
> of the vendors are worth checking. I have them bookmarked out of
> professional interest ;-)
>

I did find FCC references but those seemed ambiguous as well. Thanks for
the wi-fi.org link.

Should I send a patch renaming SDIO_DEVICE_ID_BROADCOM_CYPRESS_43752 ->
SDIO_DEVICE_ID_BROADCOM_43752 ?

Thanks
Angus

> Regards,
> Arend

2022-12-03 17:26:17

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On December 3, 2022 4:37:19 PM Angus Ainslie <[email protected]> wrote:

> On 2022-12-03 07:22, Arend Van Spriel wrote:
>> On December 3, 2022 2:56:39 PM Angus Ainslie <[email protected]> wrote:
>>
>>> On 2022-11-29 11:48, Arend van Spriel wrote:
>>>> On 11/17/2022 7:36 PM, Angus Ainslie wrote:
>>>>> On 2022-11-17 10:12, Arend Van Spriel wrote:
>>>>>> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]>
>>>>>> wrote:
>>>>>>
>>>>>>> On 2022-11-17 04:01, Andy Green wrote:
>>>>>>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>>>>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>>>>>>> Got a new version of downstream driver and firmware here:
>>>>>>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>>>>>>
>>>>>>>>> Thanks for that. The firmware indeed seems to originate from
>>>>>>>>> Synaptics
>>>>>>>>> and this module can be found on wifi-alliance product finder:
>>>>>>>>>
>>>>>>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>>>>>>
>>>>>>>>> We have no known contacts there and they do not seem to have
>>>>>>>>> much
>>>>>>>>> interest in open-source. In that gitlab repo I see they are
>>>>>>>>> providing
>>>>>>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>>>>>>
>>>>>>>>> Will see what I can do. First have to think on how to approach
>>>>>>>>> this
>>>>>>>>> failure to get most out of a debug effort. I do not have the
>>>>>>>>> module
>>>>>>>>> over here so we would need to go back and forth. The fact that
>>>>>>>>> the
>>>>>>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>>>>>>
>>>>>>>> Yes it seems that at some point with some firmware, SDIO has
>>>>>>>> worked
>>>>>>>> for, eg, Angus. And, it should work well with bchdhd (but that
>>>>>>>> does
>>>>>>>> not seem to work for scan for me on v6.1-rc5 uplevel with the
>>>>>>>> firmware
>>>>>>>> load changed to use kernel apis; but the firmware has no timeout
>>>>>>>> on
>>>>>>>> bcmdhd). So it doesn't feel a million miles away.
>>>>>>>
>>>>>>> This is the firmware we have been using for SDIO. It was sent to
>>>>>>> us
>>>>>>> directly by SparkLAN so not sure how it compares to other
>>>>>>> variants.
>>>>>>>
>>>>>>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>>>>>>
>>>>>> It is a tad bit older:
>>>>>>
>>>>>> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
>>>>>> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
>>>>>> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
>>>>>> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>>>>>
>>>>> Currently it's broken for hotspot mode but I haven't had the chance
>>>>> to
>>>>> see if that is FW or something further up the stack.
>>>>
>>>> Hi Angus,
>>>>
>>>> When you added support for this device in brcmfmac you marked it as a
>>>> Cypress device. Are you sure about that? AFAICT this device is coming
>>>> from Synaptics.
>>>
>>> Hi Arend,
>>>
>>> I assumed Cypress based on the fact that it most closely resembled the
>>> 4373 from the driver side.
>>
>> Right. Both are based on the same chip architecture that they brought
>> from Broadcom so the similarities are not surprising.
>>
>>> How would I figure out whether it's Cypress or Synaptics ?
>>>
>>> When I search the web for it I only find SparkLAN or "AMPAK
>>> technology"
>>> references.
>>
>> Earlier my colleague Franky found this link to WFA certificate:
>>
>> https://api.cert.wi-fi.org/api/certificate/download/public?variantId=103641
>>
>> WFA and FCC are usually good sources of info. Also the product pages
>> of the vendors are worth checking. I have them bookmarked out of
>> professional interest ;-)
>
> I did find FCC references but those seemed ambiguous as well. Thanks for
> the wi-fi.org link.
>
> Should I send a patch renaming SDIO_DEVICE_ID_BROADCOM_CYPRESS_43752 ->
> SDIO_DEVICE_ID_BROADCOM_43752 ?

Feel free to do so. I have some cleanup work queued up so I can also take
care of it.

Regards,
Arend



Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature

2022-12-04 14:42:30

by Angus Ainslie

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On 2022-12-03 09:08, Arend Van Spriel wrote:
> On December 3, 2022 4:37:19 PM Angus Ainslie <[email protected]> wrote:
>
>> On 2022-12-03 07:22, Arend Van Spriel wrote:
>>> On December 3, 2022 2:56:39 PM Angus Ainslie <[email protected]> wrote:
>>>
>>>> On 2022-11-29 11:48, Arend van Spriel wrote:
>>>>> On 11/17/2022 7:36 PM, Angus Ainslie wrote:
>>>>>> On 2022-11-17 10:12, Arend Van Spriel wrote:
>>>>>>> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> On 2022-11-17 04:01, Andy Green wrote:
>>>>>>>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>>>>>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>>>>>>>> Got a new version of downstream driver and firmware here:
>>>>>>>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>>>>>>>
>>>>>>>>>> Thanks for that. The firmware indeed seems to originate from
>>>>>>>>>> Synaptics
>>>>>>>>>> and this module can be found on wifi-alliance product finder:
>>>>>>>>>>
>>>>>>>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>>>>>>>
>>>>>>>>>> We have no known contacts there and they do not seem to have
>>>>>>>>>> much
>>>>>>>>>> interest in open-source. In that gitlab repo I see they are
>>>>>>>>>> providing
>>>>>>>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>>>>>>>
>>>>>>>>>> Will see what I can do. First have to think on how to approach
>>>>>>>>>> this
>>>>>>>>>> failure to get most out of a debug effort. I do not have the
>>>>>>>>>> module
>>>>>>>>>> over here so we would need to go back and forth. The fact that
>>>>>>>>>> the
>>>>>>>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>>>>>>>
>>>>>>>>> Yes it seems that at some point with some firmware, SDIO has
>>>>>>>>> worked
>>>>>>>>> for, eg, Angus. And, it should work well with bchdhd (but that
>>>>>>>>> does
>>>>>>>>> not seem to work for scan for me on v6.1-rc5 uplevel with the
>>>>>>>>> firmware
>>>>>>>>> load changed to use kernel apis; but the firmware has no
>>>>>>>>> timeout
>>>>>>>>> on
>>>>>>>>> bcmdhd). So it doesn't feel a million miles away.
>>>>>>>>
>>>>>>>> This is the firmware we have been using for SDIO. It was sent to
>>>>>>>> us
>>>>>>>> directly by SparkLAN so not sure how it compares to other
>>>>>>>> variants.
>>>>>>>>
>>>>>>>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>>>>>>>
>>>>>>> It is a tad bit older:
>>>>>>>
>>>>>>> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
>>>>>>> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
>>>>>>> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
>>>>>>> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>>>>>>
>>>>>> Currently it's broken for hotspot mode but I haven't had the
>>>>>> chance
>>>>>> to
>>>>>> see if that is FW or something further up the stack.
>>>>>
>>>>> Hi Angus,
>>>>>
>>>>> When you added support for this device in brcmfmac you marked it as
>>>>> a
>>>>> Cypress device. Are you sure about that? AFAICT this device is
>>>>> coming
>>>>> from Synaptics.
>>>>
>>>> Hi Arend,
>>>>
>>>> I assumed Cypress based on the fact that it most closely resembled
>>>> the
>>>> 4373 from the driver side.
>>>
>>> Right. Both are based on the same chip architecture that they brought
>>> from Broadcom so the similarities are not surprising.
>>>
>>>> How would I figure out whether it's Cypress or Synaptics ?
>>>>
>>>> When I search the web for it I only find SparkLAN or "AMPAK
>>>> technology"
>>>> references.
>>>
>>> Earlier my colleague Franky found this link to WFA certificate:
>>>
>>> https://api.cert.wi-fi.org/api/certificate/download/public?variantId=103641
>>>
>>> WFA and FCC are usually good sources of info. Also the product pages
>>> of the vendors are worth checking. I have them bookmarked out of
>>> professional interest ;-)
>>
>> I did find FCC references but those seemed ambiguous as well. Thanks
>> for
>> the wi-fi.org link.
>>
>> Should I send a patch renaming SDIO_DEVICE_ID_BROADCOM_CYPRESS_43752
>> ->
>> SDIO_DEVICE_ID_BROADCOM_43752 ?
>
> Feel free to do so. I have some cleanup work queued up so I can also
> take care of it.
>

If you're already going to be in there it would be great if you could
clean it up.

Thanks
Angus

> Regards,
> Arend

2022-12-04 15:41:25

by Arend van Spriel

[permalink] [raw]
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf

On December 4, 2022 3:33:02 PM Angus Ainslie <[email protected]> wrote:

> On 2022-12-03 09:08, Arend Van Spriel wrote:
>> On December 3, 2022 4:37:19 PM Angus Ainslie <[email protected]> wrote:
>>
>>> On 2022-12-03 07:22, Arend Van Spriel wrote:
>>>> On December 3, 2022 2:56:39 PM Angus Ainslie <[email protected]> wrote:
>>>>
>>>>> On 2022-11-29 11:48, Arend van Spriel wrote:
>>>>>> On 11/17/2022 7:36 PM, Angus Ainslie wrote:
>>>>>>> On 2022-11-17 10:12, Arend Van Spriel wrote:
>>>>>>>> On November 17, 2022 5:06:25 PM Angus Ainslie <[email protected]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> On 2022-11-17 04:01, Andy Green wrote:
>>>>>>>>>> On 11/17/22 09:23, Arend van Spriel wrote:
>>>>>>>>>>> On 11/16/2022 1:25 PM, Shengyu Qu wrote:
>>>>>>>>>>>> Got a new version of downstream driver and firmware here:
>>>>>>>>>>>> https://gitlab.com/rk3588_linux/linux/external/rkwifibt/-/tree/linux-5.10-gen-rkr3.4
>>>>>>>>>>>
>>>>>>>>>>> Thanks for that. The firmware indeed seems to originate from
>>>>>>>>>>> Synaptics
>>>>>>>>>>> and this module can be found on wifi-alliance product finder:
>>>>>>>>>>>
>>>>>>>>>>> https://www.wi-fi.org/product-finder-results?sort_by=certified&sort_order=desc&keywords=ap6275
>>>>>>>>>>>
>>>>>>>>>>> We have no known contacts there and they do not seem to have
>>>>>>>>>>> much
>>>>>>>>>>> interest in open-source. In that gitlab repo I see they are
>>>>>>>>>>> providing
>>>>>>>>>>> bcmdhd driver and I assume they consider that sufficient.
>>>>>>>>>>>
>>>>>>>>>>> Will see what I can do. First have to think on how to approach
>>>>>>>>>>> this
>>>>>>>>>>> failure to get most out of a debug effort. I do not have the
>>>>>>>>>>> module
>>>>>>>>>>> over here so we would need to go back and forth. The fact that
>>>>>>>>>>> the
>>>>>>>>>>> issue is seen on both SDIO and PCIe is somewhat encouraging.
>>>>>>>>>>
>>>>>>>>>> Yes it seems that at some point with some firmware, SDIO has
>>>>>>>>>> worked
>>>>>>>>>> for, eg, Angus. And, it should work well with bchdhd (but that
>>>>>>>>>> does
>>>>>>>>>> not seem to work for scan for me on v6.1-rc5 uplevel with the
>>>>>>>>>> firmware
>>>>>>>>>> load changed to use kernel apis; but the firmware has no
>>>>>>>>>> timeout
>>>>>>>>>> on
>>>>>>>>>> bcmdhd). So it doesn't feel a million miles away.
>>>>>>>>>
>>>>>>>>> This is the firmware we have been using for SDIO. It was sent to
>>>>>>>>> us
>>>>>>>>> directly by SparkLAN so not sure how it compares to other
>>>>>>>>> variants.
>>>>>>>>>
>>>>>>>>> https://source.puri.sm/Librem5/firmware-brcm43752-nonfree/-/tree/master/firmware/brcm
>>>>>>>>
>>>>>>>> It is a tad bit older:
>>>>>>>>
>>>>>>>> $ strings brcmfmac43752-sdio.bin | grep FW-TAG
>>>>>>>> <FW-TAG>4362a2-roml/config_sdio_ampak_release_he
>>>>>>>> Version=18.35.387.23.7 ([email protected]) (wlan=r880297)
>>>>>>>> Date=2020-07-13T10:34:00Z FWID=01-f50fdff0
>>>>>>>
>>>>>>> Currently it's broken for hotspot mode but I haven't had the
>>>>>>> chance
>>>>>>> to
>>>>>>> see if that is FW or something further up the stack.
>>>>>>
>>>>>> Hi Angus,
>>>>>>
>>>>>> When you added support for this device in brcmfmac you marked it as
>>>>>> a
>>>>>> Cypress device. Are you sure about that? AFAICT this device is
>>>>>> coming
>>>>>> from Synaptics.
>>>>>
>>>>> Hi Arend,
>>>>>
>>>>> I assumed Cypress based on the fact that it most closely resembled
>>>>> the
>>>>> 4373 from the driver side.
>>>>
>>>> Right. Both are based on the same chip architecture that they brought
>>>> from Broadcom so the similarities are not surprising.
>>>>
>>>>> How would I figure out whether it's Cypress or Synaptics ?
>>>>>
>>>>> When I search the web for it I only find SparkLAN or "AMPAK
>>>>> technology"
>>>>> references.
>>>>
>>>> Earlier my colleague Franky found this link to WFA certificate:
>>>>
>>>> https://api.cert.wi-fi.org/api/certificate/download/public?variantId=103641
>>>>
>>>> WFA and FCC are usually good sources of info. Also the product pages
>>>> of the vendors are worth checking. I have them bookmarked out of
>>>> professional interest ;-)
>>>
>>> I did find FCC references but those seemed ambiguous as well. Thanks
>>> for
>>> the wi-fi.org link.
>>>
>>> Should I send a patch renaming SDIO_DEVICE_ID_BROADCOM_CYPRESS_43752
>>> ->
>>> SDIO_DEVICE_ID_BROADCOM_43752 ?
>>
>> Feel free to do so. I have some cleanup work queued up so I can also
>> take care of it.
>
> If you're already going to be in there it would be great if you could
> clean it up.

Sure. No problem.

Regards,
Arend



Attachments:
smime.p7s (4.12 kB)
S/MIME Cryptographic Signature