On 07.06.12 17:04:26, Greg KH wrote: > On Thu, Jun 07, 2012 at 09:49:01AM +0200, Borislav Petkov wrote: > > On Thu, Jun 07, 2012 at 10:35:04AM +0900, Greg KH wrote: > > > On Wed, Jun 06, 2012 at 05:13:29PM +0200, Robert Richter wrote: > > > > Minimal stable backport of f227d4306cf30e1d5b6f231e8ef9006c34f3d186. > > > > > > Minimal doesn't begin to describe this. A totally different patch is > > > more like it. > > > > > > Why wasn't it done this way originally? I really don't like diverging > > > from Linus's tree in large ways like this. > > > > > > Borislav, what do you think about this version? Any reason why your > > > original patch shouldn't just be applied to 3.4 instead? > > > > Right, so Robert thought the original patch is too big for stable, > > that's why. But the main change is lvt_interrupt_supported() while the > > rest are contextual changes to accomodate that first thing. And I don't > > think it is too big, so if you were to ask me, I'd backport the original > > patch which is already upstream and widely tested... > > I'd prefer to take the original patch as well. > > Oh, and the other thing is, this patch indirectly fixes IBS registration > > so if you want to use IBS with perf on -stable, you need that patch and > > I didn't say that in the commit message. And this is actually the main > > reason why we want it in stable - to enable precise instruction tracing > > on AMD. > > What patch would this be that is needed also?
It is either f227d4306cf30e1d5b6f231e8ef9006c34f3d186 or this stable one. I am fine with both. Both patches effectivly run setup_APIC_mce() with the lvt offset from bank 4 which fixes the IBS setup issue. There is a small conflict with the original patch version and at least v3.0, but this can easily be solved. Ideally the original patch would have been splitted in a stable fix and a new implementation. But I missed this too while reviewing it. Thanks, -Robert -- Advanced Micro Devices, Inc. Operating System Research Center -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
