some updates: On 8/15/05, Dror Cohen <dror.xiv@xxxxxxxxx> wrote: > Hello all, >.... > when using the latest 6.1.16 version of the driver (which doesn't seem > to have the watch dog fix, but a lot of other fixes) we see similar > results to our modified version - there are still NETDEV events with > the "Detected Tx Unit Hang" message followed by the output of the > registers. we just had a stuck machine with the 6.1.16 version (again, no console , had to reboot) var log messages looks like the one the previous halt (order allocation failed, from e1000 irq handling). I guess this means the backporting of the watch dog does indeed help. > > more information can be supplied if needed. > > our questions: > * Is anyone experiencing similar problems (looks to us alot like: "[ > 955064 ] e1000 freezes Linux with 82544GC (similar to a year-old bug)") ? forgot the link: http://sourceforge.net/tracker/index.php?func=detail&aid=955064&group_id=42302&atid=447449 > > * Was the watch dog change forgotten, or left out deliberetly ? Does our > patch seems to be OK ? it is also not present in the latest version (6.1.6). > > * are the allocation problems related in anyway to e1000 problem ? > > ... > > > since I'm not on the linux-kernel list (but I am on the linux-net and e1000 lists) please CC replies to dror at xiv.co.il thanks, Dror - : send the line "unsubscribe linux-net" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
