> would be nice to have some debugging apparatus for bugs like this. > Perhaps artificially add a large pfn, then convert/unconvert, then > subtract it and expect the whole transformation to be an identity > mapping? This way we could simulate most of the effects of >4GB RAM, > right?
If you want to really test it regularly hacking KVM or Xen or Qemu to supply large virtual addresses is probably better. Long ago we tested with a special BIOS that always put some memory high up the map. But then Addresses >3GB are pretty common now because a lot of people put in 2 2 GB DIMMs into systems (2GB DIMMs are not really that expensive anymore). And with memory hole remapping you get suitable large addresses then. So it should be tested often enough. On the other hand the whole area was always subtly broken. I yesterday finally tracked down ftp://ftp.firstfloor.org/pub/ak/x86_64/quilt/patches/pgtable-nx which was broken forever. What an embarassing bug :/ -Andi -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
