I have been having some curious problem with rtl8187 for about a week
- having problem associating with 'direct probe to AP <mac-addrr>
timed out' after a reboot, and/or switching between compat-wireless
modules. I thought it is a temporary problem with
wireless-testing/compat-wireless but obviously nobody else has seen
anything like mine from the mailing-list traffic. So I tried
bisecting, etc and the results are inconsistent. Then I came to a
vaguely working solution - it seems that I cannot modprobe
compat-wireless twice - i.e. I can associate with compat-wireless
after make install, make unload, modprobe; but modprobe -r then
modprobe -v again I can't associate . I thought about restarting
network manager, wpa_supplicant, etc without success, but the surest
way to re-gain connectivity is simply to blow away /lib/modules/`uname
-r`/updates, depmod -a and load the as-shipped modules. (and then I
can make install compat-wireless and use it, and the cycle repeats if
I unload compat-wireless and try to install a different
compat-wireless snapshot, etc).
This is really curious, as it seems that I cannot unload
compat-wireless and reload it (or a later snapshot) without going
through as-shipped-by-distro modules ( 2.6.29.4-167.fc11.x86_64) in
the middle. How is that possible? One possibility is that as-shipped
module does a reset of hardware on unload or wireless-testing put the
hardware in a bad state on unload; but I know of no change of that
sort. Anybody else having problems associating after modprobe -r
/modprobe -v with recent compat-wireless/wireless-testing?
2009/5/31 G?bor Stefanik <[email protected]>:
> I've been experiencing something similar in vmware, though I thought
> it was vmware's fault. (I have an 8187L.) In my case, detaching and
> re-attaching the device to the VM was sufficient to reset it.
>
How long have you had this problem? Mine is integrated internally to
the laptop (although logically being on "usb") so physically detaching
the device is not possible. I sync up with compat-wireless every few
days and my problem seem to start about a week ago (during John's
travel? and lack of much commit-activity between 23th and 27th).
I just suspend'ed/resume'd with compat-wireless alright (that unloads
rtl8187, and also obviously does a lot of other things) so some
sequence of unload/reload works, just not manually modprobe -r/-v a
2nd time after make install +make unload +modprobe.
The only thing that changed recently enough is this - but the date is
still too old; unless it is the old git thing where the date is on the
local/private repository and only pushed to wireless-testing recently
- I have no idea how to check when a commit first appears in wireless
testing, vs what the date says?:
commit 16cf1f65c90cc47f6c21ca22b4accfa9749b4977
Author: Johannes Berg <[email protected]>
Date: Thu Apr 23 16:13:26 2009 +0200
mac80211: unify config_interface and bss_info_changed
The config_interface method is a little strange, it contains the
BSSID and beacon updates, while bss_info_changed contains most
other BSS information for each interface. This patch removes
config_interface and rolls all the information it previously
passed to drivers into bss_info_changed.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
I've been experiencing something similar in vmware, though I thought
it was vmware's fault. (I have an 8187L.) In my case, detaching and
re-attaching the device to the VM was sufficient to reset it.
On Sun, May 31, 2009 at 6:56 AM, Hin-Tak Leung <[email protected]> wrote:
> I have been having some curious problem with rtl8187 for about a week
> - having problem associating with 'direct probe to AP <mac-addrr>
> timed out' after a reboot, and/or switching between compat-wireless
> modules. I thought it is a temporary problem with
> wireless-testing/compat-wireless but obviously nobody else has seen
> anything like mine from the mailing-list traffic. So I tried
> bisecting, etc and the results are inconsistent. Then I came to a
> vaguely working solution - it seems that I cannot modprobe
> compat-wireless twice - i.e. I can associate with compat-wireless
> after make install, make unload, modprobe; but modprobe -r then
> modprobe -v again I can't associate . I thought about restarting
> network manager, wpa_supplicant, etc without success, but the surest
> way to re-gain connectivity is simply to blow away /lib/modules/`uname
> -r`/updates, depmod -a and load the as-shipped modules. (and then I
> can make install compat-wireless and use it, and the cycle repeats if
> I unload compat-wireless and try to install a different
> compat-wireless snapshot, etc).
>
> This is really curious, as it seems that I cannot unload
> compat-wireless and reload it (or a later snapshot) without going
> through as-shipped-by-distro modules ( 2.6.29.4-167.fc11.x86_64) in
> the middle. How is that possible? One possibility is that as-shipped
> module does a reset of hardware on unload or wireless-testing put the
> hardware in a bad state on unload; but I know of no change of that
> sort. Anybody else having problems associating after modprobe -r
> /modprobe ?-v with recent compat-wireless/wireless-testing?
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to [email protected]
> More majordomo info at ?http://vger.kernel.org/majordomo-info.html
>
--
Vista: [V]iruses, [I]ntruders, [S]pyware, [T]rojans and [A]dware. :-)