Date: Sat, 29 Oct 2011 00:25:40 -0700 From: matt <sendtomatt@gmail.com> To: =?ISO-8859-1?Q?Gustau_P=E9rez?= <gperez@entel.upc.edu> Cc: freebsd-wireless@freebsd.org Subject: Re: RT2860/RT3090/RAL: Any brave testers? Message-ID: <4EABAA74.2090409@gmail.com> In-Reply-To: <4EA18789.5070502@entel.upc.edu>
index | next in thread | previous in thread | raw e-mail
On 10/21/11 07:54, Gustau Pérez wrote: >> I've been working on combing ral & Alexandr Egorenkov's rt2860. It runs >> fine here (LiteOn v0x1814 d0x3090), and compiles sans complaint with >> both clang and gcc. >> >> However, I haven't tested any other rt2860/3090 devices, or traditional >> ral devices (they should be the same, changes only to allow rt2860 attach). >> >> If anyone has a device with a ralink chipset, and wants to test a >> development quality wireless driver :), please let know. >> >> Diffs against CVS >> /usr/src/sys/modules/ral/Makefile: http://pastebin.com/cgHy3SXL >> /usr/src/sys/dev/ral: http://pastebin.com/wXHd7fG9 >> >> Please let me know if there is better way to distribute or if you'd like >> a tarball. There are still some spurious printfs on attach about >> HT/WifiN...I think they can be ignored. >> >> This is not a production driver. Please use for testing only. There is >> definitely work to be done. If you test it, and it works for you (esp. >> LEDs, HT, injection, hostap etc.) let me know. >> >> Thanks! >> Matt > > Hi, > > I tried to use the code with FreeBSD 10.0-CURRENT r226440. My > system is an Asus eeePC 1000H with a Ralink RT2860, this is the device > info: > > ral0@pci0:1:0:0: class=0x028000 card=0x27901814 > chip=0x07811814 rev=0x00 hdr=0x00 > vendor = 'Ralink corp.' > device = 'RT2860' > class = network > > I had to manually edit sys/dev/ral/if_ral_pci.c, because the card > is identified as a RT2790. I also had to remove device ral from my > kernel config file because it couldn't be linked with the kernel. > > After rebooting the machine, I kldloaded the module by hand. I got > an "The invalid EEPROM LNA gain" and the "could not find extension > channel" errors. I saw a previous post telling us not to worry too > much about them, but I have to say they still appear. This is the > exact trace that I got: > > ral0: <Ralink Technology RT2860> mem 0xfbef0000-0xfbefffff irq > 19 at device 0.0 on pci1 > ral0: invalid EEPROM LNA gain #2: 0x00 > ral0: invalid EEPROM LNA gain #3: 0x00 > ral0: invalid EEPROM powersave level > ral0: MAC/BBP RT2860 (rev 0x28720200), RF RT3022 2.4G 2T2R > ral0: skip channel 10, could not find extension channel > ral0: skip channel 11, could not find extension channel > ral0: skip channel 12, could not find extension channel > ral0: skip channel 13, could not find extension channel > ral0: skip channel 14, could not find extension channel > > I cannot associate with any AP. I changed those AP to ensure they > were not using any of the channels in the range of 10..14 (those > channels are the ones ral0 complains with an skip channel message). I > can't even do an scan, ifconfig wlan0 list scan up returns an empty list. > > How can I proceed further? Can I provide you any additional > information? Thanks Gustau! For some reason I sometimes had to do ifconfig wlan0 up && sleep 2 && ifconfig wlan0 scan sometimes. As though it needed a moment of initialization to scan. You also have a different RF than I have, so anything's possible. Unfortunately I've been very busy lately, but I can assure you that anything that happens with ral probably won't use that code anymore. I've been working on and off in a direct port from OpenBSD. It's become clear that either the rt2860 mentioned by Michael or OpenBSD's ral or (both if they are the same?) will probably be the path. I won't get to even look at OpenBSD's ral again until a week or so from now, but I do intend to give it another shot, I just wouldn't hold your breath. My recommendation is to try the code offered by Michael, it sounds as though it is working. Matthome | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4EABAA74.2090409>
