Message-ID: <83877838-1203-4e21-9612-bc43407eaa54@collabora.com>
Date: Tue, 14 Oct 2025 11:20:44 +0100
From: Martyn Welch <martyn.welch@...labora.com>
To: Jeff Chen <jeff.chen_1@....com>
Cc: "linux-wireless@...r.kernel.org" <linux-wireless@...r.kernel.org>,
 "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
 "briannorris@...omium.org" <briannorris@...omium.org>,
 "johannes@...solutions.net" <johannes@...solutions.net>,
 "francesco@...cini.it" <francesco@...cini.it>,
 Pete Hsieh <tsung-hsien.hsieh@....com>,
 "s.hauer@...gutronix.de" <s.hauer@...gutronix.de>,
 Brian Hsu <brian.hsu@....com>
Subject: Re: [EXT] Re: [PATCH v5 00/22] wifi: nxpwifi: create nxpwifi to
 support iw61x



On 13/10/2025 19:18, Jeff Chen wrote:
> Hi Martyn,
> 
> Thanks again for your detailed feedback and testing.
> 
>  From what you've described, this issue doesn't appear to be caused by the
> nxpwifi driver itself. The Wi-Fi side seems to be functioning correctly.
> 
> To simplify things and avoid firmware loading conflicts, I recommend using
> the Wi-Fi-only firmware with the nxpwifi driver, and letting the btnxpuart
> driver handle its own Bluetooth firmware loading.
> Also, I've looped in Neeraj and Amit, who are the maintainer of the btnxpuart
> driver.
> 
>  From your BT-only scenario logs:
> We can confirm that the host-to-chip communication is working correctly, and the
> firmware download completes successfully. This rules out hardware issues and
> confirms that the UART link is functional. The problem seems to occur during HCI initialization,
> where the btnxpuart driver reports repeated Frame reassembly failed (-84) errors.
> 
> We suggest that to debug this quickly, it would be very helpful if you could capture the UART
> lines (TX/RX/CTS/RTS) using a logic analyzer (e.g., Saleae) and share the trace. This would allow
> us to inspect what's happening on the wire and identify the root cause of the
> "Frame Reassembly Failed" error. There are affordable logic analyzers compatible with Saleae
> software, such as:
> https://www.amazon.in/Robodo-Electronics-USBLA24M-Analyser-Compatible/dp/B07B8KHYDD?source=ps-sl-shoppingads-lpcontext&ref_=fplfs&psc=1&smid=AJ6SIZC8YQDZX
> 

Hi Jeff,

Thanks for the information.

I have a logic analyser, but those lines are running from BGA to a PCB 
module that's 95% canned and currently I have no access to them.

Unfortunately I can't risk breaking this device trying to access them.

Martyn

> best regards,
> 
> Jeff
>> -----Original Message-----
>> From: Martyn Welch <martyn.welch@...labora.com>
>> Sent: Saturday, October 11, 2025 12:22 AM
>>
>> On 09/10/2025 17:01, Jeff Chen wrote:
>>> On Tue, Oct 07, 2025 at 05:49:59 PM +0100, Martyn Welch wrote:
>>>> This seems to be working well for me, at least for Wifi. I'd like to
>>>> get Bluetooth up as well. The bluetooth driver (btnxpuart) doesn't
>>>> seem happy loading the bluetooh firmware:
>>>
>>
>> Yes, I'm currently using the wifi only firmware. If I'm not mistaken, that's the
>> only firmware that the NXPWifi driver will currently use. It appears that it
>> should be able to load `firmware_sdiouart`, but that doesn't appear to be set
>> anywhere and thus it's not considered. I tried to rename the combo firmware
>> to have the name used for the wifi only firmware, but that was failing for me
>> (more about that in a moment).
>>
>>> To help us reproduce and investigate further, could you please share:
>>> - Your test platform (SoC, board, etc.)
>>
>> Custom board based on Renesas RZ-G2L, using Ezurio Sona NX611 M.2 1216
>> module.
>>
>> Kernel is 6.17 with an added custom DT, a few patches for other hardware and
>> the NXPWifi patch series.
>>
>> Both btnxpuart and nxpwifi are built as modules.
>>
>> The firmware itself is what Ezurio provide in their Yocto layer:
>>
>> $ md5sum *
>> 6a9307d27c3bdb3bde800265056ab217  sd_w61x_v1.bin.se
>> 8a28ec7f1b77dbde0ac7568d0426c669  sduart_nw61x_v1.bin.se
>> d38935f03dbe6da7a9ac3daf58e640bf  uartspi_n61x_v1.bin.se
>>
>>
>> Looking a bit deeper, if I try and load the combined firmware, the bluetooth
>> modules attempts to load before the WiFi driver has had a chance to complete
>> loading the firmware which is resulting in the firmware failing.
>>
>> If I blacklist the btnxpuart module (and rename the combo firmware), it loads:
>>
>>> - Your test steps
>>
>> At the moment, booting and checking to see if the hardware has probed.
>> If it has, ensure I can connect to my wifi network and ensure that I can pass a
>> little bit of traffic.
>>
>>> - The exact firmware version you're using
>>
>> As above.
>>
>>> Also, have you tried running Bluetooth in a BT-only scenario (without
>>> loading the Wi-Fi driver)? Does that work correctly?
>>>
>>
>> I hadn't. Just gave that a go and I seem to be getting the same failure (though :
>>
>> $ sudo dmesg | grep hci0
>> [   40.859055] Bluetooth: hci0: ChipID: 7601, Version: 0
>> [   40.889487] Bluetooth: hci0: Request Firmware:
>> nxp/uartspi_n61x_v1.bin.se
>> [   42.850682] Bluetooth: hci0: FW Download Complete: 396444 bytes
>> [   42.850736] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   44.222243] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   44.230226] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   44.237480] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   44.247567] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   44.254188] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   44.263459] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   44.269814] Bluetooth: hci0: Frame reassembly failed (-84)
>> [   46.280494] Bluetooth: hci0: Opcode 0x1002 failed: -110
>> [   46.285835] Bluetooth: hci0: command 0x1002 tx timeout
>>
> 
>> Are there any BT patches that I'm missing?
>>
>> Martyn
>>
>>> Thanks,
>>> Jeff
> 

Powered by blists - more mailing lists