Closed francescamen closed 3 years ago
Thank for debugging this, I think you are right and this is just a printing error. Were you able to see any actual change in behaviour when you set the peer antenna mask like this?
I added the fix to my 5.10 and higher drivers, easy enough to add to older ones if someone cares to. It is cosmetic issue anyway.
I'm trying to use the functionality to set tx-antenna pattern on a per-peer basis. My hardware has 4 antennas and I want to assign two of them to one specific peer with id 1. As stated in the instructions, if I want to set peer-id 0x1 to use antenna pattern 0xc I have to do the following:
# echo 0x130001000c > ct_special
However, when I check if the set has been executed correctly, looking at the dmesg output, I see the following:ath10k_pci 0000:01:00.0: Setting ct-andmask for peer: 1 to 0x4.
I don't know if the problem is related to hardware constraints or it is only a problem with the print. In that respect, I'm wondering why, in the file debug.c, the value to be printed is masked with 0x16. I would mask it with 0x1F instead.ath10k_warn(ar, "Setting ct-andmask for peer: %d to 0x%x.\n", val >> 16, val & 0x16);
Is this the problem? Thank you.Hardware: Netgear R7800 NIC chipset: QCA9984 Firmware version 10.4b-ct