Sorry for jumping in a bit late... On Tuesday, 7 Aug, Andrea Arcangeli wrote: > On Wed, Aug 01, 2001 at 06:53:17PM -0400, Daniel Potts wrote: > > It appears to me that the self-mapping entry in each task page table is > > somehow getting corrupted, causing user mode permission bits to be set. I > > don't even know why we have this mapping because a running kernel does not > > use it. While the patch does make it go away, I'm still concerned that > > something just happened to trash that memory. This one needs to be examined > > closer. > > > > Line 63 of linux-2.4.7/arch/alpha/mm/init.c > > I think the problem is that that is an entry in the pgd, it's not a pte, > so it's a kind of corrupted entry in first place. I also don't know how > can it be useful so I like your patch.
I always thought that the self-mapping entry was used by PALcode to quickly handle TLB misses. When the appropriate entries of page table self-mapping are not cached in DTBs, PALcode has to descend through page tables. I don't remember what happens if the self-mapping is not provided at all - if things are still working, it means that PALcode has to go the "slow" way for every TLB miss. Nikita
