Messages in this thread | | | Date | Tue, 28 Sep 2010 11:38:54 -0400 | From | Don Zickus <> | Subject | Re: [PATCH -v2 3/7] x86, NMI, Rename memory parity error to PCI SERR error |
| |
On Tue, Sep 28, 2010 at 09:33:37AM +0800, Huang Ying wrote: > On Tue, 2010-09-28 at 00:45 +0800, Robert Richter wrote: > > > > > Ok. I will add CPU ID in message. Because we know the reason, I don't > > > > > think we need the reason in message. > > > > > > > > You only know that bit 7 is set, not the rest. As this is an error > > > > message we should provide as much information as possible. > > > > > > Well, what other info do we know besides that bit being set? (I wish we > > > had more, but I don't think we do) > > > > We should keep printing the reason byte as it did before. > > The reason is printed before because mem_parity_error is treated as > something like unknown reason. And iochk_error is treated as known > reason and will not print the reason byte. Please the check the original > code. > > But now we treat pci_serr_error (renamed from mem_parity_error) as known > reason. So it is not necessary to print the reason byte. I suggest to > print the reason byte only if (!(reason & 0xc0) && reason), where the > reason is really unknown.
So you are just matching the iochk_error. I understand your reasoning. Yesterday I thought you changed the unknown_nmi_error code. I guess I don't see it as a problem, but I'll let Robert chime in.
Cheers, Don
| |