On Mon, Aug 25, 2014 at 11:20 PM, Emmanuel Grumbach <egrumbach@xxxxxxxxx> wrote: >> Hi Christian, >> >> I'm not the best person to talk about the iwlwifi firmwares (that would >> be Emmanuel, but he is not available now), but I'll try. > > Here I am - thanks Luca for stepping in. > >> >> On Sat, 2014-08-23 at 00:01 +0200, Christian Lamparter wrote: >>> Hello, >>> >>> I got new hardware. Among the new stuff, there was one >>> Intel® Dual Band Wireless-AC 3160. When I tried to connect >>> to one (in fact any) of my networks, I was getting the >>> following errors: >>> >>> [ 1021.092378] iwlwifi 0000:01:00.0: Microcode SW error detected. Restarting 0x82000000. >>> [...] >>> [ 1021.093935] iwlwifi 0000:01:00.0: Start IWL Error Log Dump: >>> [ 1021.093939] iwlwifi 0000:01:00.0: Status: 0x00000000, count: 6 >>> [ 1021.093942] iwlwifi 0000:01:00.0: Loaded firmware version: 25.223.9.0 >>> [ 1021.093947] iwlwifi 0000:01:00.0: 0x00000038 | BAD_COMMAND <--- >>> [ 1021.093951] iwlwifi 0000:01:00.0: 0x00A00220 | uPc >> >> Where did you get this firmware version? AFAICT the last version >> Emmanuel sent to linux-firmware.git is 23.214.9.0 (the one you also >> mentioned below). >> > > That's the Core6 FW that I "published" in a side branch in my > linux-firmware clone. > >> >>> [...] >>> [ 1021.094192] Call Trace: >>> [ 1021.094205] [<ffffffff813ce9bd>] ? dump_stack+0x41/0x51 >>> [ 1021.094217] [<ffffffffa0b9cd8b>] ? iwl_trans_pcie_send_hcmd+0x367/0x438 [iwlwifi] >>> [ 1021.094226] [<ffffffffa0b9b52c>] ? iwl_pcie_enqueue_hcmd+0x5e3/0x5fe [iwlwifi] >>> [ 1021.094235] [<ffffffff8105fd4c>] ? finish_wait+0x5d/0x5d >>> [ 1021.094250] [<ffffffffa0d50106>] ? iwl_mvm_send_cmd+0xd/0x22 [iwlmvm] >>> [ 1021.094262] [<ffffffffa0d50160>] ? iwl_mvm_send_cmd_pdu+0x45/0x52 [iwlmvm] >>> [ 1021.094273] [<ffffffffa0d4c5fb>] ? iwl_mvm_configure_bcast_filter.isra.14.part.15+0x36/0x3f [iwlmvm] >>> >>> (Full log attached below) >>> >>> What cough my eye was the "BAD_COMMAND" in the error log. I disabled >>> "CONFIG_IWLWIFI_BCAST_FILTERING" option and the 3160 is now working. >>> >>> As far as I can tell, there is already a firmware feature flag: >>> "IWL_UCODE_TLV_FLAGS_BCAST_FILTERING" which can be used to >>> distinguish whenever the feature is supported or not. But it >>> seems to be "stuck on" for the 3160. >> >> Weird, but it's possible that they set the flag mistakenly for the 3160 >> firmware. I'll check this with the firmware team. >> > > Right - we need to check with the firmware team (I already asked - but > things take time). > I'd say that the firmware mistakenly advertises support for this > feature. I bet that 3160 doesn't support it, but that's yet to be > checked. > Actually, we do have an internal ticket with this issue - but I don't > know if it was on 3160 or on 7260. Your report explicitly says that > you saw this on 3160, so I am only now connecting between the dots. I > can't check the internal bug since I am on vacation right now. Note > that 3160 is much less common than 7260 - so this kind of stupid bugs > can stay unseen for a fairly long time... > > >> >>> Regards >>> Christian >>> >>> BTW: I also tried the older "23.214.9.0" firmware too. >>> => Same result [Doesn't work with BCAST_FILTERING enabled]. >> >> Thanks a lot for your report! Broadcast filtering is not set by default, >> so I haven't seen that bug before. One more thing - Chrisian, please open a bug on bug as described here: http://wireless.kernel.org/en/users/Drivers/iwlwifi#Bugzilla This can help to track things. Thanks. -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
