On 02/12/2012 09:10 PM, Aaron Z wrote:
> On 02/02/2012 4:52:09 PM, Hauke Mehrtens wrote:
>> On 01/28/2012 06:55 AM, Aaron Z wrote:
>>> When I run "killall -9 hostapd" then "wifi" I get told:
>>> "Configuration file: /var/run/hostapd-phy0.conf" and "Using
>>> interface wlan0 with hwaddr MACADDRESS and ssid 'OpenWrt'"wifi is
>>> still rejecting connections.
>>> When I run "killall -9 hostapd; /usr/sbin/hostapd -dd
>>> /var/run/hostapd-phy0.conf > /tmp/hostapd.log" it works right
>>> away.
>> This behavior is strange it could be that debugging activates some
>> needed code path, hostapd just works after the second start or that
>> there are timing issues. Could you run hostapd just with -d and
>> without
>> -d and report what happens in these cases?
> Sorry for the delay. Its been busy around here.
> I ran "killall -9 hostapd; /usr/sbin/hostapd -d /var/run/hostapd-phy0.conf > 
> /tmp/hostapd.log" I was able to connect and go online. The log is attached as 
> hostapd -d.txt (4 KB).
> I then rebooted the WAP and ran "killall -9 hostapd; /usr/sbin/hostapd 
> /var/run/hostapd-phy0.conf > /tmp/hostapd.log" I was able to connect and go 
> online. The logfile was blank, so I didn't attach it.
> 
> I tried connecting before running the commands and was rejected both times.
> 
> Thanks
> 
> Aaron Z
Hi Aaron Z,

have you set up a build environment? If so please try the attached patch
( just place it under package/mac80211/patches/) and build b43 with
debug enabled. I had some problems with n-phy device on bcma as well and
it was fixed with this patch. If the problem still exists please post
the output of dmesg after the first boot.

Hauke
>From cd6e7f84fa813c48f608f95152f020c842cc8e5a Mon Sep 17 00:00:00 2001
From: Hauke Mehrtens <ha...@hauke-m.de>
Date: Sat, 18 Feb 2012 22:21:01 +0100
Subject: [PATCH 202/202] b43: do not use msleep(1) in b43_radio_init2056_post

With msleep(1) I get a Data bus error and with usleep_range(1500, 2000)
it works. I udelay(1000) also works, but usleep_range(1000, 1500) does
not work. The timer documentation says you should not use msleep for
values < 20.

Signed-off-by: Hauke Mehrtens <ha...@hauke-m.de>
---
 drivers/net/wireless/b43/phy_n.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

--- a/drivers/net/wireless/b43/phy_n.c
+++ b/drivers/net/wireless/b43/phy_n.c
@@ -661,7 +661,7 @@ static void b43_radio_init2056_post(stru
 	b43_radio_set(dev, B2056_SYN_COM_CTRL, 0xB);
 	b43_radio_set(dev, B2056_SYN_COM_PU, 0x2);
 	b43_radio_set(dev, B2056_SYN_COM_RESET, 0x2);
-	msleep(1);
+	udelay(1000);
 	b43_radio_mask(dev, B2056_SYN_COM_RESET, ~0x2);
 	b43_radio_mask(dev, B2056_SYN_PLL_MAST2, ~0xFC);
 	b43_radio_mask(dev, B2056_SYN_RCCAL_CTRL0, ~0x1);
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel

Reply via email to