Re: WCN6855: Request: Support ThinkPad X13s 21BY-S03Y00^[

From: Johan Hovold <johan@kernel.org>
To: Alex R <alex@ironrobin.net>
Cc: Jeff Johnson <quic_jjohnson@quicinc.com>,
	Kalle Valo <kvalo@kernel.org>,
	"ath11k@lists.infradead.org" <ath11k@lists.infradead.org>
Subject: Re: WCN6855: Request: Support ThinkPad X13s 21BY-S03Y00^[
Date: Mon, 11 Sep 2023 13:26:56 +0200	[thread overview]
Message-ID: <ZP75gMQVajZIJlIL@hovoldconsulting.com> (raw)
In-Reply-To: <aMi4zXfWxMWs2QxBRyLwLb30U-nf6DPgkOd27if32rD-T5rYfBcS5q8fXgUE-FAarK7NlWmNjDt12o81fDVLJMtygm5mpracZ2TuvrVoFQg=@ironrobin.net>

On Thu, Sep 07, 2023 at 03:16:15PM +0000, Alex R wrote:
> Q. Did this module ever work, and if so with which kernel version?
> A. It seems to work more reliably on my archiso kernel, which is 5.19.0. 
> 
> IIRC this is where we were using the wifi hack, before we got the new
> X13s board file.

Yes, but it seems you never get as far as loading the board file so I
don't think that one's involved here.

> Q. How is this module identified when it starts successfully? 
> 
> A. Here is the dmesg from a "successful" bootup: https://pastebin.com/1kLDzXSD
> (I was able to get the wlan0 device to come up by booting into Windows
> 11, checking for updates, and then shutting down (not rebooting) 
> 
> (fast boot Windows 11 setting is disabled)

Ok, so you do have a different chip_id than I do (0x2 instead of 0x12),
but both ids have an entry in board-2.bin (otherwise it would not work
"sometimes"):

	ath11k_pci 0006:01:00.0: chip_id 0x2 chip_family 0xb board_id 0x8c soc_id 0x400c0210

You could compare it with what the chip in your other machine which does
not have this issue is identified as (compare all the values above).

> Q. How did things change in 6.5?
> A. On both 6.4 and 6.5 of those versions, we have the same issue in
> the dmesg logs. 
> 
> Although it seems I am able to get "successful" boots more often with 6.4.
> I have no data on this though and could be wrong.

Ok, got it. But with 5.19 you never saw this issue at all?

> Q. Have you verified that you're using the same fw on both the working
> and non-working machine (not just board-2.bin, which does not seem to
> be involved here)?
> A. I use my own distro based on Arch Linux ARM project, and use the
> same x13s-firmware package on both machines, so firmware must be the
> same.
> https://github.com/ironrobin/x13s-alarm/tree/trunk/x13s-firmware

Ok, the fw version is also logged on successful probe and you appear to
use the same version as I do:

	ath11k_pci 0006:01:00.0: fw_version 0x110b196e fw_build_timestamp 2022-12-22 12:54 fw_build_id WLAN.HSP.1.1-03125-QCAHSPSWPL_V1_V2_SILICONZ_LITE-3.6510.23

Could you check what fw version you were using with your 5.19 iso image?

> Q. You also said something on IRC about dual-booting Windows making a
> difference, etc. etc.
> A. On my faulty machine, I've tried both dual booting Windows and
> running linux as the only root partition on a nvme drive.
> The only difference I've noticed is that I can more reliably bring
> back the wlan0 device by booting into Windows 11, connecting to the
> internet, and then rebooting into Linux.

Hmm... Not sure what's going on here, and I'm afraid I won't have time
to look at this myself for a while.

Johan

-- 
ath11k mailing list
ath11k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath11k