Messages in this thread | | | Date | Thu, 2 Dec 2010 15:37:03 -0800 (PST) | From | David Rientjes <> | Subject | Re: [patch 7/7, v7] NUMA Hotplug Emulator: Implement mem_hotplug/add_memory debugfs interface |
| |
On Thu, 2 Dec 2010, shaohui.zheng@intel.com wrote:
> From: Shaohui Zheng <shaohui.zheng@intel.com> > > Add mem_hotplug/add_memory interface to support to memory hotplug emulation. > the reserved memory can be added into desired node with this interface. > > Add a memory section(128M) to node 3(boots with mem=1024m) > > echo 0x40000000,3 > mem_hotplug/add_memory > > And more we make it friendly, it is possible to add memory to do > > echo 3g > mem_hotplug/add_memory > echo 1024m,3 > mem_hotplug/add_memory > > Another format suggested by Dave Hansen: > > echo physical_address=0x40000000 numa_node=3 > mem_hotplug/add_memory > > it is more explicit to show meaning of the parameters. >
NACK, we don't need such convoluted definitions if debugfs were extended with per-node triggers to add_memory as I suggested in v6 of your proposal:
/sys/kernel/debug/mem_hotplug/add_node (already exists) /sys/kernel/debug/mem_hotplug/node0/add_memory /sys/kernel/debug/mem_hotplug/node1/add_memory ...
You can then write a physical starting address to the add_memory files to hotadd memory to a node other than the one to which it has physical affinity. This is much more extendable if we add additional per-node triggers later.
It would also be helpful if you were to reach consensus on the matters under discussion before posting a new version of your patchset everyday.
| |