glouis@dynamicro.on.ca wrote: > > Last night I installed 2.4.1-ac4 remotely on two machines that had been > running -ac3. Neither was essential to production so I rebooted > remotely to try it. One of them works fine. The other hangs at boot > with an interrupt synch problem (sorry, but I'm away from the office and > have only sketchy secondhand info; will reproduce Friday and supply > details if needed). The contents of /proc/scsi/aic7xxx/0 follow for > each (the machine that failed has been rebooted with -ac3). > =========================== > 2.4.1-ac4 fails to access disk, said to be an "interrupt synch > problem," 2.4.1-ac3 running: > > Adaptec AIC7xxx driver version: 5.2.1/5.2.0 > Compile Options: > TCQ Enabled By Default : Disabled > AIC7XXX_PROC_STATS : Enabled > > Adapter Configuration: > SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter > Ultra Wide Controller at PCI 0/11/0 There was a minor bug in the aic7xxx.seq script (actually, it was more like I failed to take into account a limitation of the parsing program that parses the binary sequencer code). It resulting in invalid instructions getting downloaded to every controller prior to a 7895. The attached patch should solve people's problems with this. (Note, this patch wasn't made against a full linux tree, so cd into driver/scsi and use patch -p1 in order to apply it). -- Doug Ledford http://people.redhat.com/dledford Please check my web site for aic7xxx updates/answers before e-mailing me about problems