Bug 14808
Summary: | rt2800usb does not work, rt2870sta does | ||
---|---|---|---|
Product: | Drivers | Reporter: | Josip Rodin (joy+kernel) |
Component: | network-wireless | Assignee: | Gertjan van Wingerde (gwingerde) |
Status: | CLOSED PATCH_ALREADY_AVAILABLE | ||
Severity: | normal | CC: | andreas.sturmlechner, bas, gwingerde, IvDoorn, linville, spsppl |
Priority: | P1 | ||
Hardware: | x86-64 | ||
OS: | Linux | ||
Kernel Version: | 2.6.32 | Subsystem: | |
Regression: | No | Bisected commit-id: |
Description
Josip Rodin
2009-12-14 13:09:03 UTC
Could you provide the dmesg output of the non-working driver? By the looks of it you are having an rt3070 based device, which indeed isn't working yet in rt2x00. Proper support for those devices hopefully can reach the upstream kernel in 2.6.34. I removed the staging module, made sure nothing in lsmod mentioned rt*, and did a modprobe rt2800usb, and this is what syslog says: Dec 14 18:39:14 pork kernel: cfg80211: Calling CRDA to update world regulatory domain Dec 14 18:39:14 pork kernel: phy0: Selected rate control algorithm 'minstrel' Dec 14 18:39:14 pork kernel: usbcore: registered new interface driver rt2800usb Dec 14 18:39:14 pork kernel: udev: renamed network interface wlan0 to wlan1 Dec 14 18:39:14 pork kernel: rt2800usb 1-9:1.0: firmware: requesting rt2870.bin Dec 14 18:39:14 pork kernel: ADDRCONF(NETDEV_UP): wlan1: link is not ready Dec 14 18:39:16 pork wpa_supplicant[3581]: Failed to initiate AP scan. Dec 14 18:39:16 pork dhclient: Internet Systems Consortium DHCP Client V3.1.3 Dec 14 18:39:16 pork dhclient: Copyright 2004-2009 Internet Systems Consortium. Dec 14 18:39:16 pork dhclient: All rights reserved. Dec 14 18:39:16 pork dhclient: For info, please visit https://www.isc.org/software/dhcp/ Dec 14 18:39:16 pork dhclient: Dec 14 18:39:17 pork wpa_supplicant[3581]: CTRL-EVENT-SCAN-RESULTS Dec 14 18:39:17 pork wpa_supplicant[3581]: Trying to associate with 00:21:04:4b:2c:a0 (SSID='ShadowGreen' freq=2417 MHz) Dec 14 18:39:17 pork kernel: wlan1: deauthenticating from 00:21:04:4b:2c:a0 by local choice (reason=3) Dec 14 18:39:17 pork kernel: wlan1: direct probe to AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:17 pork kernel: wlan1: direct probe responded Dec 14 18:39:17 pork kernel: wlan1: authenticate with AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:17 pork kernel: wlan1: authenticated Dec 14 18:39:17 pork kernel: wlan1: associate with AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:17 pork kernel: wlan1: RX AssocResp from 00:21:04:4b:2c:a0 (capab=0x431 status=0 aid=4) Dec 14 18:39:17 pork kernel: wlan1: associated Dec 14 18:39:17 pork wpa_supplicant[3581]: Associated with 00:21:04:4b:2c:a0 Dec 14 18:39:17 pork kernel: ADDRCONF(NETDEV_CHANGE): wlan1: link becomes ready Dec 14 18:39:17 pork dhclient: Listening on LPF/wlan1/00:1f:1f:55:8c:a3 Dec 14 18:39:17 pork dhclient: Sending on LPF/wlan1/00:1f:1f:55:8c:a3 Dec 14 18:39:17 pork dhclient: Sending on Socket/fallback Dec 14 18:39:19 pork dhclient: DHCPDISCOVER on wlan1 to 255.255.255.255 port 67 interval 3 Dec 14 18:39:20 pork kernel: wlan1: deauthenticated from 00:21:04:4b:2c:a0 (Reason: 2) Dec 14 18:39:20 pork wpa_supplicant[3581]: WPA: 4-Way Handshake failed - pre-shared key may be incorrect Dec 14 18:39:20 pork wpa_supplicant[3581]: CTRL-EVENT-DISCONNECTED - Disconnect event - remove keys Dec 14 18:39:21 pork wpa_supplicant[3581]: CTRL-EVENT-SCAN-RESULTS Dec 14 18:39:21 pork wpa_supplicant[3581]: Trying to associate with 00:21:04:4b:2c:a0 (SSID='ShadowGreen' freq=2417 MHz) Dec 14 18:39:21 pork kernel: wlan1: direct probe to AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:21 pork kernel: wlan1: direct probe responded Dec 14 18:39:21 pork kernel: wlan1: authenticate with AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:21 pork kernel: wlan1: authenticated Dec 14 18:39:21 pork kernel: wlan1: associate with AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:21 pork kernel: wlan1: RX ReassocResp from 00:21:04:4b:2c:a0 (capab=0x431 status=0 aid=4) Dec 14 18:39:21 pork kernel: wlan1: associated Dec 14 18:39:21 pork wpa_supplicant[3581]: Associated with 00:21:04:4b:2c:a0 Dec 14 18:39:22 pork dhclient: DHCPDISCOVER on wlan1 to 255.255.255.255 port 67 interval 7 Dec 14 18:39:24 pork kernel: wlan1: deauthenticated from 00:21:04:4b:2c:a0 (Reason: 2) Dec 14 18:39:24 pork wpa_supplicant[3581]: WPA: 4-Way Handshake failed - pre-shared key may be incorrect Dec 14 18:39:24 pork wpa_supplicant[3581]: CTRL-EVENT-DISCONNECTED - Disconnect event - remove keys Dec 14 18:39:25 pork wpa_supplicant[3581]: CTRL-EVENT-SCAN-RESULTS Dec 14 18:39:25 pork wpa_supplicant[3581]: Trying to associate with 00:21:04:4b:2c:a0 (SSID='ShadowGreen' freq=2417 MHz) Dec 14 18:39:25 pork kernel: wlan1: direct probe to AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:25 pork kernel: wlan1: direct probe responded Dec 14 18:39:25 pork kernel: wlan1: authenticate with AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:25 pork kernel: wlan1: authenticated Dec 14 18:39:25 pork kernel: wlan1: associate with AP 00:21:04:4b:2c:a0 (try 1) Dec 14 18:39:25 pork kernel: wlan1: RX ReassocResp from 00:21:04:4b:2c:a0 (capab=0x431 status=0 aid=4) Dec 14 18:39:25 pork kernel: wlan1: associated Dec 14 18:39:25 pork wpa_supplicant[3581]: Associated with 00:21:04:4b:2c:a0 Dec 14 18:39:28 pork kernel: wlan1: no IPv6 routers present Dec 14 18:39:29 pork kernel: wlan1: deauthenticated from 00:21:04:4b:2c:a0 (Reason: 2) Dec 14 18:39:29 pork dhclient: DHCPDISCOVER on wlan1 to 255.255.255.255 port 67 interval 8 Dec 14 18:39:29 pork wpa_supplicant[3581]: WPA: 4-Way Handshake failed - pre-shared key may be incorrect Dec 14 18:39:29 pork wpa_supplicant[3581]: CTRL-EVENT-DISCONNECTED - Disconnect event - remove keys Dec 14 18:39:30 pork wpa_supplicant[3581]: CTRL-EVENT-SCAN-RESULTS Dec 14 18:39:30 pork wpa_supplicant[3581]: Trying to associate with 00:21:04:4b:2c:a0 (SSID='ShadowGreen' freq=2417 MHz) ... Repetitio ad nauseam. I'm noticing the explicit errors now - but they're still wrong. The wlan1 wpa_supplicant configuration file is the same one the working (staging) driver uses, I literally reference the exact same file from interfaces(5). OK. The part that I'm after doesn't show up in syslog. Can you post the output of dmesg (Just type dmesg on the command line). I'm mostly interested in lines like this: [ 6.088988] phy3 -> rt2x00_set_chip: Info - Chipset detected - rt: 1600, rf: 0006, rev: 30700201. There simply isn't any of that, anywhere. Suggestions? I plugged it into another machine to eliminate the other one as a factor, and had a "d'oh" moment - it negotiates physical parameters when they - change... :) I should have just plugged it out and back in. Anyway this is what it says on the other machine, also with 2.6.32: Dec 14 19:53:40 CNZGRIJRL kernel: usb 1-1: new high speed USB device using ehci_hcd and address 4 Dec 14 19:53:40 CNZGRIJRL kernel: usb 1-1: configuration #1 chosen from 1 choice Dec 14 19:56:48 CNZGRIJRL kernel: phy2: Selected rate control algorithm 'minstrel' Dec 14 19:56:48 CNZGRIJRL kernel: Registered led device: rt2800usb-phy2::radio Dec 14 19:56:48 CNZGRIJRL kernel: Registered led device: rt2800usb-phy2::assoc Dec 14 19:56:48 CNZGRIJRL kernel: Registered led device: rt2800usb-phy2::quality Dec 14 19:56:48 CNZGRIJRL kernel: usbcore: registered new interface driver rt2800usb Dec 14 19:58:48 CNZGRIJRL kernel: rt2800usb 1-1:1.0: firmware: requesting rt2870.bin Dec 14 19:58:48 CNZGRIJRL kernel: ADDRCONF(NETDEV_UP): wlan1: link is not ready It gets stuck there, as this machine's wpa_supplicant doesn't do anything with it - the link remains down even though I'm actually much closer to the AP on this machine... the green indicator light on the card just keeps blinking several times a second. I'll try it on the original machine in a bit. OK. BTW, you will need to enable CONFIG_RT2X00_DEBUG to get the message I am after. Oh, okay. Here we go: usb 2-2: new high speed USB device using ehci_hcd and address 5 usb 2-2: configuration #1 chosen from 1 choice phy4 -> rt2x00_set_chip: Info - Chipset detected - rt: 1600, rf: 0005, rev: 30700201. phy4: Selected rate control algorithm 'minstrel' Registered led device: rt2800usb-phy4::radio Registered led device: rt2800usb-phy4::assoc Registered led device: rt2800usb-phy4::quality usbcore: registered new interface driver rt2800usb phy4 -> rt2x00lib_request_firmware: Info - Loading firmware file 'rt2870.bin'. rt2800usb 2-2:1.0: firmware: requesting rt2870.bin phy4 -> rt2x00lib_request_firmware: Info - Firmware detected - version: 0.10. phy4 -> rt2x00mac_conf_tx: Info - Configured TX queue 0 - CWmin: 3, CWmax: 4, Aifs: 2, TXop: 102. phy4 -> rt2x00mac_conf_tx: Info - Configured TX queue 1 - CWmin: 4, CWmax: 5, Aifs: 2, TXop: 188. phy4 -> rt2x00mac_conf_tx: Info - Configured TX queue 2 - CWmin: 5, CWmax: 10, Aifs: 3, TXop: 0. phy4 -> rt2x00mac_conf_tx: Info - Configured TX queue 3 - CWmin: 5, CWmax: 10, Aifs: 7, TXop: 0. ADDRCONF(NETDEV_UP): wlan1: link is not ready (In reply to comment #7) > phy4 -> rt2x00_set_chip: Info - Chipset detected - rt: 1600, rf: 0005, rev: > 30700201. OK. That explains it. It indeed is a RT3070 device which unfortunately isn't supported yet. Stay tuned, I may have some patches available in the next few weeks for this. Thanks. Isn't this a bit confusing to have a device supported by the upstream 28* driver while it's not actually 28* :) fwiw, I have the same problems with my sitecom usb device (0df6:003f). The chipset is slightly different though: phy0 -> rt2x00_print_chip: Info - Chipset detected - rt: 1800, rf: 0006, rev: 30700201. I tried with today's git, but although to seems to be detected and initialized fine, it simply doesn't return any scanning results. This is rather confusing, actually. Would it be possible to show a warning when a device is detected that is not yet supported? *** Bug 15300 has been marked as a duplicate of this bug. *** *** Bug 15535 has been marked as a duplicate of this bug. *** 2.6.36 should support these devices properly now. Can any of you test that kernel version to see that your devices work okay? I'm on 2.6.35-23-generic and its been working for awhile now. I have a few different Alfa cards and iwconfig is reporting the correct driver in all cases. OK. Then I'll close this bug entry. |