Sharyathi Nagesh wrote:
Dave
I wanted to bring to notice two more issue that I am currently working.
One regarding changes done with kmem_cache to make it NUMA compliant.
'lists' which was earlier pointing to kmem_list3 structure, lists has
been renamed first as nodelists than it is made into an array of pointer
to kmem_list3 structures. So now we have to traverse the whole list
iteratively to get the same information.
As a result of these change kmem -s command fails. To accomodate
these changes we need to change the function do_slab_chain_percpu_v2 to
access nodelists iteratively. Still working on it to figure out how I
can accommodate these changes.
Hi Sharyathi,
Excellent! I really appreciate your undertaking of this task.
Anyway, having gone through this kind of "keeping-up-with-the-slab-subsystem"
task many times, I would suggest that you move it into a new function rather
than re-working do_slab_chain_percpu_v2(). Even if there ends up being a bit
a bit of code duplication, future maintenance will be far easier if you make it
a new function. I would guess that development will be easier too, and you
won't have to worry about breaking backwards-compatibility.
Second objectsize in kmem_cache has been changed to object_size and this
is enabled only on selecting config option CONFIG_DEBUG_SLAB in the
kernel.
That's true, but hasn't the functionality of the old "objsize" been
replaced
by the new "buffer_size"?
Anyway, again, I appreciate this very much. I had been planning to address
this, but have been sidetracked totally looking at xen kernel issues.
Thanks,
Dave