memcg: page cgroup diet

Here is v4. I removed RFC and fixed a fatal bug in v3.

This patch series is a preparetion for removing page_cgroup->flags. To remove
it, we need to reduce flags on page_cgroup. In this set, PCG_MOVE_LOCK and
PCG_FILE_MAPPED are remvoed.

After this, we only have 3 bits.
==
enum {
        /* flags for mem_cgroup */
        PCG_LOCK,  /* Lock for pc->mem_cgroup and following bits. */
        PCG_USED, /* this object is in use. */
        PCG_MIGRATION, /* under page migration */
        __NR_PCG_FLAGS,
};
==

I'll make a try to merge this 3bits to lower 3bits of pointer to
memcg. Then, we can make page_cgroup 8(4)bytes per page.


Major changes since v3 are
 - replaced move_lock_page_cgroup() with move_lock_mem_cgroup().
   passes pointer to memcg rather than page_cgroup.

Working on linux-next feb13 and passed several tests.

Thanks,
-Kame

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/