On Sun, 2011-01-09 at 16:52 -0500, John David Anglin wrote: > On Wed, 22 Dec 2010, James Bottomley wrote: > > The kernel has an 8M tmpailas space (originally designed for copying > > and clearing pages but now only used for clearing). The idea is > > to place zeros into the cache above a physical page rather than into > > the physical page and flush the cache, because often the zeros end up > > being replaced quickly anyway. > > > > We can also use the tmpalias space for flushing a page. The difference > > here is that we have to do tmpalias processing in the non access data and > > instruction traps. The principle is the same: as long as we know the physical > > address and have a virtual address congruent to the real one, the flush will > > be effective. > > > > In order to use the tmpalias space, the icache miss path has to be enhanced to > > check for the alias region to make the fic instruction effective. > > Since I began testing this change, I have started seeing problems with the > console input on gsyprf11 and my rp3440. If I HUP getty, the console works > for awhile and then stops working again. Suspect the "INEQIVALENT ALIAS" > messages somehow kill the console. I get a lot of these doing make check > for GCC from the tests that intentionally generate a segv to test exception > support. In theory, the patch is actually stricter than our original flushing support because it eliminates the ineffective flushes and detects inequvalent aliases (and flushes all inequivalent addresses). > I have also had at least one console related HPMC. Analysis is attached. > I don't fully understand the actual cause of the HPMC (buffer overrun?). > The kernel was built with GCC 4.5.3. The faulting instruction appears > to have the base and index interchanged, although this shouldn't affect > linux. I thought this issue was fixed as a fair bit of work on this was > done in the middle-end. > > I have to say that gsyprf11 running a SMP kernel built with 4.5.3 is > more stable with your change than it has been for a long time. So I'm not sure what to make of this ... any memory fault shouldn't result in a HPMC ... I probably need to decode the HPMC to see what its saying ... I'll get around to that on the Weekend, hopefully. James -- To unsubscribe from this list: send the line "unsubscribe linux-parisc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html