Well, the kernel giveth and the kernel taketh away, or something like that. I've installed 11.04 and hadn't had any major glitches until I suspended/resumed a machine with an RTL8192SU plugged in. The adapter would not resume. If I unplugged and plugged in again it works perfectly
Googling shows this bug: https://lists.linux-foundation.org/p...ry/026559.html so I don't feel so lonely.
Relevant parts of dmesg after a suspend/resume cycle
[44820.162843] PM: resume of drv:sd dev:2:0:0:0 complete after 5573.547 msecs
[44820.162884] PM: resume of drv:scsi_disk dev:2:0:0:0 complete after 2546.730 msecs
[44820.162891] PM: resume of drv:scsi_device dev:2:0:0:0 complete after 5573.567 msecs
[44820.162942] PM: resume of devices complete after 5578.383 msecs
[44820.167933] usblp0: USB Bidirectional printer dev 6 if 0 alt 0 proto 2 vid 0x03F0 pid 0x1604
[44820.168092] PM: resume devices took 5.584 seconds
[44820.168121] PM: Finishing wakeup.
[44820.168122] Restarting tasks ... done.
[44820.252043] r8169 0000:02:00.0: eth0: link down
[44820.252408] ADDRCONF(NETDEV_UP): eth0: link is not ready
[44820.361033] ADDRCONF(NETDEV_UP): wlan1: link is not ready
[44820.432232] r8712u: in r8711_wx_set_scan: bDriverStopped=1
[44820.749316] EXT4-fs (sda1): re-mounted. Opts: errors=remount-ro,commit=0
[44821.002176] r8712u: in r8711_wx_set_scan: bDriverStopped=1
[44841.003106] r8712u: in r8711_wx_set_scan: bDriverStopped=1
After unplugging then replugging the adapter
[45049.543789] usb 1-1: USB disconnect, address 7
[45051.812142] usb 1-1: new high speed USB device using ehci_hcd and address 8
[45051.947763] r8712u: DriverVersion: v7_0.20100831
[45051.947803] r8712u: register rtl8712_netdev_ops to netdev_ops
[45051.947811] r8712u: USB_SPEED_HIGH with 4 endpoints
[45051.949085] r8712u: Boot from EFUSE: Autoload OK
[45052.686170] r8712u: CustomerID = 0x0000
[45052.686180] r8712u: MAC Address from efuse = 00:14:d1:e7:17:ba
[45052.721366] <30>udev: renamed network interface wlan0 to wlan1
[45053.444618] r8712u: 1 RCR=0x153f00e
[45053.445365] r8712u: 2 RCR=0x553f00e
[45053.553108] ADDRCONF(NETDEV_UP): wlan1: link is not ready
[45060.845353] r8712u: [r8712_got_addbareq_event_callback] mac = 00:26:62:7c:b1:5b, seq = 0, tid = 7
[45060.845684] ADDRCONF(NETDEV_CHANGE): wlan1: link becomes ready
[45063.737587] r8712u: [r8712_got_addbareq_event_callback] mac = 00:26:62:7c:b1:5b, seq = 0, tid = 0
[45071.744072] wlan1: no IPv6 routers present
[45075.632704] r8712u: [r8712_got_addbareq_event_callback] mac = 00:26:62:7c:b1:5b, seq = 512, tid = 0
This device works perfectly in Maverick which uses the 2.6.35 kernel when the firmware patch is applied. I assume that firmware patch won't work where kernel level support is present. Any thoughts about how to work around this problem until this bug is squashed?
Edit: When I run lsmod I see what appear to be two realtek modules loaded, r8169 and r8712u. I was thinking dueling modules so I blacklisted r8169. Restart and r8169 still showed up in lsmod and the device worked after restart. I then went back and blacklisted r8712u and restarted. The device didn't come alive and r8712u was not listed by lsmod. Obviously the devices uses r8712u. I wonder why r8169 loads after being blacklisted when r8712u does not?