Linux wireless drivers development
 help / color / mirror / Atom feed
From: Andy Green <andy@warmcat.com>
To: Shengyu Qu <wiagn233@outlook.com>, arend.vanspriel@broadcom.com
Cc: franky.lin@broadcom.com, linux-wireless@vger.kernel.org
Subject: Re: AP6275 / bcm43752 pcie on mainline brcmf
Date: Sun, 13 Nov 2022 09:41:19 +0000	[thread overview]
Message-ID: <de7f4df1-54b0-db2e-3d40-3caeda020924@warmcat.com> (raw)
In-Reply-To: <TY3P286MB26116261C165BD4B0B249EB098029@TY3P286MB2611.JPNP286.PROD.OUTLOOK.COM>



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 <andy@warmcat.com> 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

  reply	other threads:[~2022-11-13  9:41 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1D540F2B-17E8-4992-BD06-4AA4A66B1964@warmcat.com>
     [not found] ` <1845da4b838.279b.9b12b7fc0a3841636cfb5e919b41b954@broadcom.com>
     [not found]   ` <b5f564da-d0c1-4617-fef1-771ccdda18e4@warmcat.com>
2022-11-10  7:06     ` AP6275 / bcm43752 pcie on mainline brcmf Andy Green
2022-11-10  7:35       ` Arend Van Spriel
2022-11-13  8:26         ` Shengyu Qu
2022-11-13  9:03           ` Andy Green
2022-11-13  9:10             ` Shengyu Qu
2022-11-13  9:41               ` Andy Green [this message]
2022-11-13 10:04                 ` Shengyu Qu
2022-11-13 10:56             ` Arend Van Spriel
2022-11-13 15:52               ` Shengyu Qu
2022-11-13 17:35                 ` Arend Van Spriel
2022-11-13 18:43                   ` Andy Green
2022-11-16 12:25                   ` Shengyu Qu
2022-11-16 14:16                     ` Andy Green
2022-11-17  9:23                     ` Arend van Spriel
2022-11-17 12:01                       ` Andy Green
2022-11-17 16:06                         ` Angus Ainslie
2022-11-17 18:12                           ` Arend Van Spriel
2022-11-17 18:36                             ` Angus Ainslie
2022-11-29 19:48                               ` Arend van Spriel
2022-12-03 13:56                                 ` Angus Ainslie
2022-12-03 15:22                                   ` Arend Van Spriel
2022-12-03 15:37                                     ` Angus Ainslie
2022-12-03 17:08                                       ` Arend Van Spriel
2022-12-04 14:33                                         ` Angus Ainslie
2022-12-04 15:29                                           ` Arend Van Spriel
2022-11-17 21:52                         ` Arend Van Spriel
2022-11-18  5:41                           ` Andy Green
2022-11-24  9:42                             ` Arend van Spriel
2022-11-24 10:20                               ` Andy Green
2022-11-24 10:49                                 ` Arend van Spriel
2022-11-24 11:04                                   ` Andy Green
2022-11-24 11:07                                     ` Arend van Spriel
2022-11-24 11:11                                       ` Andy Green
2022-11-24 11:18                                         ` Arend van Spriel
2022-11-24 11:24                                           ` Andy Green
2022-11-24 12:14                               ` Andy Green
2022-11-24 12:22                                 ` Arend van Spriel
2022-11-24 12:24                                   ` Shengyu Qu
2022-11-24 12:26                                   ` Andy Green
2022-11-24 12:35                                     ` Shengyu Qu
2022-11-24 12:38                                       ` Arend van Spriel
2022-11-24 12:44                                     ` Arend van Spriel
2022-11-24 12:54                                       ` Andy Green
2022-11-24 13:47                                         ` Arend van Spriel
2022-11-24 14:16                                           ` Andy Green
2022-11-24 14:32                                             ` Andy Green

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=de7f4df1-54b0-db2e-3d40-3caeda020924@warmcat.com \
    --to=andy@warmcat.com \
    --cc=arend.vanspriel@broadcom.com \
    --cc=franky.lin@broadcom.com \
    --cc=linux-wireless@vger.kernel.org \
    --cc=wiagn233@outlook.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox