CVS has a fix. James G. Sack (jim) wrote: > I can report the eeprom.o module from 2.9.0 on a SuperMicro P4SCT+II > seems to produce the same bus problems reported elsewhere (intel boards, > I think). > > The symptom is repeated log messages like > PCI device 8086:25a4 (Intel Corp.): Reset failed! (01) > eeprom.o: block read fail at 0x00! > > The problem does NOT show up on a couple of Tyan boards I have > available, one a S2707 Trinity GC-SL and another of unknown ID (but it > has an ServerWorks CNB20LE northbridge and OSB4 southbridge if that's > any help) > > I tried mhoffman's patch in > http://archives.andrew.net.au/lm-sensors/msg29196.html > and it regularly hits the 'timeout >= MAX_TIMEOUT' branch. > > .jim
