Fwd: "Hardware error, resetting" on Atheros AR9285

[Adrian Chadd]

FYI, since this discussion disappeared off the list..



Adrian

---------- Forwarded message ----------
From: Adrian Chadd <adrian at freebsd.org>
Date: 23 April 2011 12:44
Subject: Re: "Hardware error, resetting" on Atheros AR9285
To: Rad Manly <radmanly at comcast.net>


Nope, unfortunately.

That sync register value is AR_INTR_SYNC_LOCAL_TIMEOUT, which
shouldn't be triggering a fatal error.

I think there may be some bugs in the interrupt handling code. I'll do
some further digging for you.

In the meantime, please apply this to sys/dev/ath:

Index: if_ath.c
===================================================================
--- if_ath.c    (revision 220950)
+++ if_ath.c    (working copy)
@@ -1265,7 +1265,7 @@
       struct ath_softc *sc = arg;
       struct ifnet *ifp = sc->sc_ifp;
       struct ath_hal *ah = sc->sc_ah;
-       HAL_INT status;
+       HAL_INT status = 0;

       if (sc->sc_invalid) {
               /*

And see if it has any effect.

Thanks,




Adrian


On 23 April 2011 11:46, Rad Manly <radmanly at comcast.net> wrote:
> OK, I'll do that.  Is there anything that immediately comes to mind as a
> possibility?  Anything I could test or check?
>
> Thanks,
> Rad
>
> On 04/22/11 21:43, Adrian Chadd wrote:
>> Please place all of this in a PR. :)
>>
>>
>> Adrian
>>
>> On 23 April 2011 08:39, Rad Manly <radmanly at comcast.net> wrote:
>>> I checked out and built the latest ath driver from -HEAD.
>>> Unfortunately, the problem persists.  I turned on all of the debugging
>>> flags and saved the log file.  Here's what I see when I load the driver:
>>>
>>> Apr 22 18:03:34 ceres kernel: ath0: <Atheros 9285> mem
>>> 0xf0100000-0xf010ffff irq 17 at device 0.0 on pci2
>>> Apr 22 18:03:34 ceres kernel: ath0: [ITHREAD]
>>> Apr 22 18:03:34 ceres kernel: [ath] AR9285E_20 detected; using XE TX
>>> gain tables
>>> Apr 22 18:03:34 ceres kernel: [ath] Enabling diversity for Kite
>>> Apr 22 18:03:34 ceres kernel: ath0: AR9285 mac 192.2 RF5133 phy 14.0
>>>
>>> Here's an example of the error:
>>>
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: 0x800965 => 0xe0001071
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: new IMR 0x8014b7
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: enable IER
>>> Apr 22 18:04:24 ceres kernel: ar5212StartTxDma: queue 1
>>> Apr 22 18:04:24 ceres kernel: ar5212StartTxDma: queue 1
>>> Apr 22 18:04:24 ceres kernel: ar5416GetPendingInterrupts: fatal error,
>>> ISR_RAC 0x0 SYNC_CAUSE 0x2000
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: 0xe0001071 => 0x0
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: disable IER
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: new IMR 0x0
>>> Apr 22 18:04:24 ceres kernel: ath0: hardware error; resetting
>>> Apr 22 18:04:24 ceres kernel: ath0: 0x00000000 0x00002000 0x00000000,
>>> 0x00000000 0x00000000 0x00000000
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: 0x0 => 0x0
>>> Apr 22 18:04:24 ceres kernel: ar5416SetInterrupts: new IMR 0x0
>>>
>>>
>>> To get this, I executed:
>>>
>>> ifconfig wlan0 create wlanmode sta wlandev ath0
>>> ifconfig wlan0 up scan
>>>
>>> Do you know what could cause this error?  I think the 0x00002000 means
>>> it's a timeout error of some kind but I don't know what's timing out or why.
>>>
>>> What can I do to help swat this bug?
>>>
>>> Thanks,
>>> Rad
>>>
>>>
>>> On 04/21/11 23:12, Adrian Chadd wrote:
>>>> Try -HEAD. I've fixed a lot of AR9285 related issues in HEAD.
>>>>
>>>>
>>>>
>>>> Adrian
>>>>
>>>> On 22 April 2011 11:14, <radmanly at comcast.net
>>>> <mailto:radmanly at comcast.net>> wrote:
>>>>
>>>>     Hello,
>>>>
>>>>     Several of us on the FreeBSD forum are having the same problem
>>>>     with the ath driver on the AR9285 chip. The driver loads fine but
>>>>     whenever I bring up an interface, I get "Hardware error,
>>>>     resetting" messages on the screen over and over again, roughly one
>>>>     a second. One user, Claude, reports he upgraded to -STABLE to no
>>>>     avail. I describe the problem in more detail in this thread:
>>>>     http://forums.freebsd.org/showthread.php?t=23297
>>>>
>>>>     Is there anything else I should try before I open a bug report?
>>>>
>>>>     Thanks,
>>>>     Rad
>>>>     _______________________________________________
>>>>     freebsd-wireless at freebsd.org <mailto:freebsd-wireless at freebsd.org>
>>>>     mailing list
>>>>     http://lists.freebsd.org/mailman/listinfo/freebsd-wireless
>>>>     To unsubscribe, send any mail to
>>>>     "freebsd-wireless-unsubscribe at freebsd.org
>>>>     <mailto:freebsd-wireless-unsubscribe at freebsd.org>"
>>>>
>>>>
>>> _______________________________________________
>>> freebsd-wireless at freebsd.org mailing list
>>> http://lists.freebsd.org/mailman/listinfo/freebsd-wireless
>>> To unsubscribe, send any mail to "freebsd-wireless-unsubscribe at freebsd.org"
>>>
>