On Tue, Jun 27, 2023 at 10:55:04AM +0200, Reyk Floeter wrote: > Hi, > > the following diff adds the Intel Wi-Fi 6 0x51f1 device, found in a > new Thinkpad X1 Carbon Gen 11. > > iwx0 at pci0 dev 20 function 3 "Intel Wi-Fi 6 AX211" rev 0x01, msix > iwx0: hw rev 0x370, fw 77.f92b5fed.0, pnvm e4a49534, address ... > > I looked at the Linux code for the values for 0x51f1 and 0x7a70: > https://github.com/torvalds/linux/commit/1c4db7613f35b248ff05b8bfb3a4580a3d11d75c
Thanks, your diff looks correct to me. ok stsp@ > This works fine with my 11ac setup and performance seems OK (around > 200Mbps - what can I expect on OpenBSD these days?). I don't know if > the device supports UHB channels. This speed matches what I see in my environment, more or less. Apparently it goes up to 300 Mbit/s for some people, but I have never seen it go that far up myself. We are capped to 80 MHz channels at the moment. I do have a 160MHz AP now but would need to find spare time to make that work (as things look right now, me finding time for this soon is rather unlikely). And the size we advertise for aggregates on Rx in VHT capabilities is the smallest possible value we can advertise in 11ac (3895 octets). Increasing this frame size limit should result in a performance boost. We might have to increase the size of our Rx buffers accordingly, unless the hardware handles de-aggregation into a cluster of our 4k-sized Rx buffers, which may well be the case but I don't know for sure. It would not be an entirely trivial diff, but likely less coding work and more widely applicable than 160 MHz channel support.
