Hello, I have set this up to be pulled into PLM automatically when you release. This way you can see the compile results on multiple architectures. Here is the current one: http://www.osdl.org/plm-cgi/plm?module=patch_info&patch_id=4410 Some of the architectures now do 'allmodconfig' rather than 'defconfig', so the hotplug options should get compiled. These are ppc64, ppc40x, x86_64, and ia64. The 'compregress' for ia32 has always done an 'allmodconfig' and 'allyesconfig' as well. Judith Lebzelter OSDL On Fri, 29 Apr 2005, Dave Hansen wrote: > Andy Whitcroft has taken back over maintaining the sparsemem patches. > I'll should be picking updates up from him in the future. He's also > planning on submitting them to the -mm tree shortly. > > The L0-sysfs-memory-class patch is no more. I'm getting it ready to be > merged, so I've cleaned up a few things: > > - removed BROKEN and INVALID memory states, we don't use these > - put a WARN() in the fall-through case(mem->state) > - remove "/* a hardware/firmware check for the ram" in > memory_probe_store() > - remove CAP_SYS_ADMIN checks. The control file permissions > should take care of this. It also leaves open the possibility > that a sysadmin could have another user/group take care of > this, instead of plain root or CAP_SYS_ADMIN. > - remove memory_hotplug_filter(). Unused, and not required to > be implemented > > Here's the set that have been broken out from that patch, or that are > closely dependent on it: > > # new infrastructure > C1-pcp_zone_init.patch > C2-globalize-flush_all_zero_pkmaps.patch > C3-__section_nr.patch > C4-zone_resize_sem.patch > # actual sysfs patches > D0-sysfs-memory-class.patch > D1-i386-hotplug-functions.patch > D2-ia64-hotplug-functions.patch > D3-x86_64-hotplug-functions.patch > # debugging only > E0-for-debugging-page_is_ram_hotplug.patch > E1-for-debugging-memory_present-for-contig-systems.patch > E2-for-debugging-handle-add-to-empty-zone.patch > E3-for-debugging-more-FLAGS_RESERVED.patch > > I'll likely submit all of these at the same time. However, I plan to > ask Andrew Morton to keep the E* patches in -mm. Those patches let you > fake memory hotplug on non-hotplug machines, but they're not horribly > useful beyond their debugging application, so they probably shouldn't go > to mainline. > > The page migration patches should be working: > > http://sr71.net/patches/2.6.12/2.6.12-rc3-mhp1/page_migration/patch-2.6.12-rc3-mhp1-pm.gz > > There are some small fix bits in here that might conflict with people's > personal trees: > > aic-compile-fix.patch > arm-pm_message_t-include.patch > ia64-acpi.patch > x86_64-rc3fix.patch > > They should back right out, so it shouldn't be much of an issue. > > Archives and broken out patches are in the same place as always: > > http://sr71.net/patches/2.6.12/2.6.12-rc3-mhp1/patch-2.6.12-rc3-mhp1.gz > http://sr71.net/patches/2.6.12/2.6.12-rc3-mhp1/ > > -- Dave > >
