Athp driver
-
Yes I have two cards in there also.
But it would be a useful test. I was referring to some of the other stuff like the Broadcom and Chelsio firmwares. I've never tried to check what is loaded there by default though. More research needed!
Steve
-
i am now on official kernel, after the introduction of ALQ, and that problem did not come out again/i don't need to modify firmware_max anymore.
also i changed my 4 lan network card from realtek to intel i350-t4. i had too many trouble from realtek.
so now i don't need to load if_re.ko anymore, maybe that helped. -
@stephenw10 i'm trying to understand why the athp0 is not recognized as a wireless interface in the gui,
i track down the problem to this file: /usr/local/www/interfaces_assign.php at line 209 where it sayif (preg_match($g['wireless_regex'], $_POST['if_add'])) { $config['interfaces'][$newifname]['wireless'] = array(); interface_sync_wireless_clones($config['interfaces'][$newifname], false); }
it should save the interface inside a 'wireless' array but it doesn't do it, it save it inside a "enable" array
i put a print_r($config); inside the php
[opt5] => Array ( [descr] => WIFI5G [if] => athp0_wlan0 [enable] => [spoofmac] => [ipaddr] => ....
vs
[opt3] => Array ( [descr] => WIFI [if] => ath0_wlan0 [wireless] => Array ( [mode] => hostap [standard] =>....from what i understand, the 'wireless' tag is readed when you go to configure the interface
i'm too tired now to go on, i will check further tomorrow, if you have any idea let me know -
Yeah I'll have to re-check tomorrow too. I'm pretty sure I found that previously though....
Steve
-
found something new inside interfaces_assign.php
if (preg_match($g['wireless_regex'] =>
/^(ath|bwi|bwn|ipw|iwi|iwm|iwn|malo|mwl|ral|rsu|rum|run|uath|upgt|ural|urtw|urtwn|wi|wpi|wtap|zyd)[0-9]
athp does not match preg_match, now i need to find out where this is defined ...found it
/etc/inc/globals.inc
need to add athp"wireless_regex" => "/^(ath|bwi|bwn|ipw|iwi|iwm|iwn|malo|mwl|ral|rsu|rum|run|uath|upgt|ural|urtw|urtwn|wi|wpi|wtap|zyd|athp)[0-9]+/",
after adding it, you need to delete and re-add the interface
-
Aha, that was it, nice!
-
Hmm, I kinda lost track of where we are with this.
I have two ath10k NICs in my test box and they load up OK across 2.5 updates. I have some kernel modules in there but they are waaay old at this point. The only line required, seemingly, is
if_athp_load="yes"
in loader.conf(.local). It seems to pull in the required firmware modules without explicitly loading them:[2.5.0-DEVELOPMENT][root@apu.stevew.lan]/root: kldstat Id Refs Address Size Name 1 22 0xffffffff80200000 3123108 kernel 2 1 0xffffffff83324000 d42f8 if_athp.ko 3 1 0xffffffff83621000 a45 athp_QCA988X_hw2.0_board.bin.ko 4 1 0xffffffff83622000 3cb9f athp_QCA988X_hw2.0_firmware-5.bin.ko 5 1 0xffffffff8365f000 fe0 cpuctl.ko 6 1 0xffffffff83660000 2678 intpm.ko 7 1 0xffffffff83663000 b10 smbus.ko 8 1 0xffffffff83664000 3110 cryptodev.ko
[2.5.0-DEVELOPMENT][root@apu.stevew.lan]/root: sysctl net.wlan.devices net.wlan.devices: athp1 athp0
I don't see any sort of driver version listed in a sysctl.
I do see it trying and failing to load this:
athp0: ath10k_fetch_fw_file: firmware_get: athp_cal-pci-athp0.bin athp_cal-pci-athp0.bin: could not load firmware image, error 2
Steve
-
hi steve,
athp_cal-pci-athp0.bin
this is a firmware for calibration, not used/needed by us. I found it month ago and searched on google for what it was.the only file we need to load are the one already present in your kldstat
the only thing that if_athp.ko do is to load the correct firmware for the card, in your case
athp_QCA988X_hw2.0_board.bin.ko and athp_QCA988X_hw2.0_firmware-5.bin.koalso i found that if you remove *firmware-5.bin.ko then firmware-4.bin.ko will be loaded.
if you remove *5.bin.ko and 4.bin.ko then *3.bin.ko will be loaded and so oni found somewhere a discussion talking about the difference of this firmware but the only thing that i rember now is that firmware-2 should be idea for AP mode. i tried all of them and saw no difference
-
Ah, interesting. Looks like the modules I have are dated May 14th, they must still be those you compiled.
I'll try to update.
Steve
-
there is nothing new from Geramy or Adrian, the last commit is 7 feb
but here is the new firmware i found
10.2.4.70.70.zip
from kvalo/ath10k-firmwarecontent of athp_QCA988X_hw2.0_firmware-5.bin.ko
^?ELF^B^A^A ...... ^@QCA-ATH10K^@w^@^@^@^@^L^@^@^@10.2.4.70.70^A^@^@^@^D^@ ......
-
Seems good:
athp1: qca988x hw2.0 (0x4100016c, 0x043202ff) fw 10.2.4.70.70 api 5 htt-ver 2.1 wmi-op 5 htt-op 2 cal otp max-sta 128 raw 0 hwcrypto 1 features no-p2p,bit10
Be nice if there was some change log. 2.5 years between those versions from what I can see.
Steve
-
nope, sorry can't find any changelog anywhere
-
Yeah I couldn't see anything either. Seems to be working OK though so no significant regression, at least not yet!
Still seeing values in scans that cannot be correct in terms of channel or even the reported MAC. Some random data in there.
Can't wait to try
media VHT mode 11ac mediaopt hostap
. Just need some way to actually apply it I guess. Fun!Steve
-
Anyone running this in a current, 12.1-stable, snapshot?
I tested the modules would load before upgrading and was surprised to find they did. But doesn't actually load the with the hardware in place.
I guess it's new modules time.
Steve
-
not anymore, I have virtualized everything on ESXi and added a Ubiquiti ap-ac-pro, for me it was too unstable.
but I have modules compiled for FreeBSD 12.1 if you want to try
https://drive.google.com/drive/folders/1fM-Jlmf8BY21kIEGueSxFWmrISZqcDj3
there is a folder athp-12.1 -
nice
Sponsored by: Rubicon Communications, LLC (d/b/a "Netgate")
Co-authored-by: Bjoern A. Zeeb bz@FreeBSD.ORG.
we had to wait for netgate to chime in before something get done
thumb up for netgate and bzfbd
but idk when we can take advantage of the patches, that drivers are intended to be built against FreeBSD 13
maybe with pfsense 2.6 ?
-
Ooo, nice I'll give them a go.
It's weird I'm seeing that firmware table full error that others reported before. It's been so long since I set this up initially I've forgotten exactly how I got past it. The notes I made that seemed sufficient at the time suddenly look woefully inadequate!Steve
-
Looks like current snaps are loading more firmware by default. Standby....
-
And we're back...
Firmware limit removed in current snaps.Steve
-
ah it was #define FIRMWARE_MAX right?