Messages in this thread | | | Date | Fri, 21 Jun 2013 10:44:06 -0600 | Subject | Re: [PATCH 0/3] iommu/amd: IOMMU Error Reporting/Handling/Filtering | From | Shuah Khan <> |
| |
On Fri, Jun 21, 2013 at 10:24 AM, Joerg Roedel <joro@8bytes.org> wrote: > On Fri, Jun 21, 2013 at 05:59:33PM +0200, Borislav Petkov wrote: >> On Fri, Jun 21, 2013 at 05:15:00PM +0200, Joerg Roedel wrote: >> > Instead of extending this bad scaling dmesg error-reporting mechanism, >> > I would very much like to see an integration of IOMMU error handling >> > into the EDAC framework. >> >> Wouldn't just adding a tracepoint and collecting the logs in debugfs be >> enough? This would at least avoid spamming syslog with the messages. > > I would rather not create another user-space interface for this which we > then have to keep forever because someone started using it. What we need > is a well defined mechanism to report these kernel metrics to userspace > which can then do with it whatever it wants. I think EDAC can provide > that interface. >
I think it makes prefect sense to use EDAC as a common frame-work for iommu errors as well. Especially if and when APEI evolves to cover IOMMU tables (both Intel VT-d and AMD) , since EDAC is already capable of working with APEI, using EDAC will work well on platforms that support APEI as well as the ones that don't.
-- Shuah Shuah Khan, Linux Kernel Developer - Open Source Group Samsung Research America (Silicon Valley) shuah.kh@samsung.com | (970) 672-0658
| |