On Thursday 27 May 2010 17:02:12 Roman Yepishev wrote: > I've been following this thread closely for some time as I have exactly > the same experience with my Acer Aspire One (AR5001 card):
it strikes me that in both cases it happens on an Acer... > 03:00.0 Ethernet controller: Atheros Communications Inc. AR5001 Wireless > Network Adapter (rev 01) Subsystem: Foxconn International, Inc. Device > e008 > Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- could you use the output of dmesg to identify the exact atheros chipset? something like "ath5k phy0: Atheros AR5414 chip found (MAC: 0xa5, PHY: 0x61)". thanks. > Usually this happens when a continuous block of data is transmitted, > i.e. during deb packages updates via http. > When corruption happens this is accompanied by > [ 469.663234] ath5k phy0: unsupported jumbo > [ 469.935929] ath5k phy0: unsupported jumbo that might be a good clue! robert do you see the same? in newer kernels we have removed the warning, however... > > for that it could be helpful to capture the whole session with wireshark > > from a third wireless card in monitor mode and see if there are any > > irregularities in the packets seen in the "air" (maybe sth. like > > retries). i know this is a lot of data and it will be hard to track the > > duplicate packet, but maybe you can find something in it, like a special > > bitpattern which you could use to search for in wireshark. > > Will try to do this as well. hmm, with WPA enabled you will not be able to match bitpatterns... > Currently it looks like something > card/vendor/revision specific thing as I have tested broadcom bc43-based > cards and two atheros (AMBIT Microsysystem / 0428) cards in the same > environment w/o any issues. so that means the other atheros cards in the acer don't have this problem, right? again please tell us the exact chipset. any chance to test the problematic card in another laptop to see if it depends on the card or the laptop/card combination? how easy is it for you to reproduce the problem? do you know any kernel version where this problem didn't happen? bruno _______________________________________________ ath5k-devel mailing list ath5k-devel@lists.ath5k.org https://lists.ath5k.org/mailman/listinfo/ath5k-devel
