Re: rtw88 import to main

From: Daniel_Pérez <>
Date: Fri, 01 Apr 2022 06:38:01 UTC
As always, thank you for your hard work, Bjoern.
Will test the mentioned stuff as soon as I am able to, hopefully I can get it built in STABLE/13.

------- Original Message -------
On Friday, April 1st, 2022 at 2:32 AM, Bjoern A. Zeeb <> wrote:

> Hi,
> I've just pushed Realtek's rtw88 driver into the tree along with
> firmware to support it and man pages. Station mode only.
> CAVEAT-0: Even though the MFC reminder is set, I'll probably not
> merge it this quickly into the stable/13 branch (see CAVEAT-1).
> If you want to do so yourself, you currently also need
> 86220d3cbd500b1018dcdabb0ba70644db438cfd if you are on i386 (I'll
> merge that in two days). I'll also keep merging anything else
> which improves things. Otherwise it's just the three commits to
> cherry-pick.
> CAVEAT-1: if you have more than 4GB of main memory, the driver will
> likely fail to load firmware or work. Sounds silly these days but hey.
> The reason for that is between LinuxKPI and busdma as I understand
> it currently and you can follow the problem by also applying
> and checking the new
> compat.linuxkpi.lkpi_pci_nseg1_fail increases to > 0 when you try
> to load the driver.
> We'll keep pondering and I'll likely post (elsewhere) more about this
> the next days. Depending on the outcome of discussions we'll see how
> to solve that problem (I am currently thinking how much else it may
> affect and whether to hack the driver or FreeBSD to fix it).
> HINT-1: If you are still curious (and given you made it here you are!)
> you can try booting with
> hw.physmem="4G" # I believe is the right tuneable
> in loader.conf (which should limit your computers ability to do anything
> interesting these days) but may make rtw88 work (and that would
> be a report I'd be very interested in along with the PCI IDs and full
> dmesg if you wish). I know from one person who has tried successfully.
> I also tried on a 2G amd64 where this worked. My 64GB arm64 is still
> not happy even with this but that's probably another story.
> SUMMARY: I've added mentions of this to
> as well.
> CAVEAT-2: because of CAVEAT-1 testing so far has been limited so
> anything you can add will be appreciated. See HINT-1.
> Happy April 1st... oh I am not supposed to say that as this is real..tek?
> /bz
> --
> Bjoern A. Zeeb r15:7