On 04/09/2014 06:44 PM, Luck, Tony wrote: >> So when the driver sees uncorrected errors, I'm also seeing them in my >> memory scanning program - so they correspond nicely. I didn't see anything >> logged in /var/log/mcelog, but I will update to the latest when possible. > I wonder if there are some BIOS options to enable reporting via CMCI/MCE? > On the E5 systems the reference BIOS uses phrases like "poison forwarding" > in the option names. > > The above behavior sounds less than useful. > > Scenario: Your mission critical app is running (controlling a giant laser > cutter). > Oops there is a memory error, and the bad data arrives at the application > causing > it to swing the laser beam through 180 degrees, destroying half of your lab. > A few seconds/minutes later - your EDAC driver prints a message saying that > the uncorrected error count just got incremented. > > -Tony
Agreed, and I don't like the polling either. But so far on this h/w I have not been able to find a better option. I also seem to recall, that ce errors tend to proceed ue errors. So, I think alerting on ce errors can help avoid getting into ue errors. So IMO there is currently value in this driver, and I know others have requested support for this h/w in the past (on the edac mailing lists). I'll see what else I can find once I get the problematic h/w in hand. Thanks, -Jason -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
