On Mon, 2018-02-12 at 19:52 +0100, Pali Rohár wrote: > On Monday 12 February 2018 20:48:47 Luca Coelho wrote: > > On Tue, 2018-01-30 at 20:49 +0100, Pali Rohár wrote: > > > Hello, > > > > Hi Pali, > > > > > > > for almost 3 years I have a problem with Intel wifi card Centrino > > > Advanced-N 6235. Its firmware periodically crash and in dmesg I'm > > > seeing > > > a message "Microcode SW error detected". > > > > > > Problem happens independently of kernel versions (I tried 3.8, > > > 3.13, > > > 3.18, 4.0, 4.9). Currently I'm using Debian Stretch and its 4.9 > > > version. > > > > > > This wifi card is identified as: > > > > > > $ lspci -s 03:00.0 -vv > > > 03:00.0 Network controller: Intel Corporation Centrino Advanced-N > > > 6235 (rev 24) > > > Subsystem: Intel Corporation Centrino Advanced-N 6235 AGN > > > > [...] > > > > > And here is example of dmesg error flood: > > > > > > [26624.998823] iwlwifi 0000:03:00.0: Microcode SW error > > > detected. Restarting 0x2000000. > > > > [...] > > > > > [26624.999013] iwlwifi 0000:03:00.0: Loaded firmware version: > > > 18.168.6.1 > > > [26624.999117] iwlwifi 0000:03:00.0: Start IWL Error Log Dump: > > > [26624.999118] iwlwifi 0000:03:00.0: Status: 0x000002CC, count: 6 > > > [26624.999120] iwlwifi 0000:03:00.0: 0x000028AA | > > > ADVANCED_SYSASSERT > > > [27464.823779] iwlwifi 0000:03:00.0: Loaded firmware version: > > > 18.168.6.1 > > > [27464.823903] iwlwifi 0000:03:00.0: Start IWL Error Log Dump: > > > [27464.823904] iwlwifi 0000:03:00.0: Status: 0x000002CC, count: 6 > > > [27464.823906] iwlwifi 0000:03:00.0: 0x000028AA | > > > ADVANCED_SYSASSERT > > > > [...] > > > > > This really looks like firmware problem. What can I do with these > > > microcode errors? Can you help me? > > > > Yes, this is a firmware problem. But unfortunately this is a > > really > > old hardware (6 years old) and we don't support the firmware > > anymore. > > Hi! That is pity :-( > > > I'm sorry, but there is really nothing we can do. > > So can I debug this problem myself? Is there any documentation or > resources for this? Nothing prevents you from trying to debug. Maybe find the exact situation when this happens and try to make a workaround in the driver. But unfortunately we don't have any public documentation for this firmware, so you'll have to try to solve it in the driver entirely. Sadly, I wouldn't have much hope in succeeding, since this is most likely a purely firmware side bug which may be impossible to work around in the driver. :( Good luck! -- Cheers, Luca.
