On Wed, Jan 24, 2024 at 07:09:19PM +0100, Uladzislau Rezki (Sony) wrote:
> This patch adds extra explanation of recently added vmap
> node layer based on community feedback. No functional change.
>
> Signed-off-by: Uladzislau Rezki (Sony) <urezki@xxxxxxxxx>
> ---
> mm/vmalloc.c | 60 ++++++++++++++++++++++++++++++++++++++++------------
> 1 file changed, 46 insertions(+), 14 deletions(-)
>
> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> index 257981e37936..b8be601b056d 100644
> --- a/mm/vmalloc.c
> +++ b/mm/vmalloc.c
> @@ -765,9 +765,10 @@ static struct rb_root free_vmap_area_root = RB_ROOT;
> static DEFINE_PER_CPU(struct vmap_area *, ne_fit_preload_node);
>
> /*
> - * An effective vmap-node logic. Users make use of nodes instead
> - * of a global heap. It allows to balance an access and mitigate
> - * contention.
> + * This structure defines a single, solid model where a list and
> + * rb-tree are part of one entity protected by the lock. Nodes are
> + * sorted in ascending order, thus for O(1) access to left/right
> + * neighbors a list is used as well as for sequential traversal.
> */
> struct rb_list {
> struct rb_root root;
> @@ -775,16 +776,23 @@ struct rb_list {
> spinlock_t lock;
> };
>
> +/*
> + * A fast size storage contains VAs up to 1M size. A pool consists
> + * of linked between each other ready to go VAs of certain sizes.
> + * An index in the pool-array corresponds to number of pages + 1.
> + */
> +#define MAX_VA_SIZE_PAGES 256
> +
> struct vmap_pool {
> struct list_head head;
> unsigned long len;
> };
>
> /*
> - * A fast size storage contains VAs up to 1M size.
> + * An effective vmap-node logic. Users make use of nodes instead
> + * of a global heap. It allows to balance an access and mitigate
> + * contention.
> */
> -#define MAX_VA_SIZE_PAGES 256
> -
> static struct vmap_node {
> /* Simple size segregated storage. */
> struct vmap_pool pool[MAX_VA_SIZE_PAGES];
> @@ -803,6 +811,11 @@ static struct vmap_node {
> unsigned long nr_purged;
> } single;
>
> +/*
> + * Initial setup consists of one single node, i.e. a balancing
> + * is fully disabled. Later on, after vmap is initialized these
> + * parameters are updated based on a system capacity.
> + */
> static struct vmap_node *vmap_nodes = &single;
> static __read_mostly unsigned int nr_vmap_nodes = 1;
> static __read_mostly unsigned int vmap_zone_size = 1;
> @@ -2048,7 +2061,12 @@ decay_va_pool_node(struct vmap_node *vn, bool full_decay)
> }
> }
>
> - /* Attach the pool back if it has been partly decayed. */
> + /*
> + * Attach the pool back if it has been partly decayed.
> + * Please note, it is supposed that nobody(other contexts)
> + * can populate the pool therefore a simple list replace
> + * operation takes place here.
> + */
> if (!full_decay && !list_empty(&tmp_list)) {
> spin_lock(&vn->pool_lock);
> list_replace_init(&tmp_list, &vn->pool[i].head);
> @@ -2257,16 +2275,14 @@ struct vmap_area *find_vmap_area(unsigned long addr)
> * An addr_to_node_id(addr) converts an address to a node index
> * where a VA is located. If VA spans several zones and passed
> * addr is not the same as va->va_start, what is not common, we
> - * may need to scan an extra nodes. See an example:
> + * may need to scan extra nodes. See an example:
> *
> - * <--va-->
> + * <----va---->
> * -|-----|-----|-----|-----|-
> * 1 2 0 1
> *
> - * VA resides in node 1 whereas it spans 1 and 2. If passed
> - * addr is within a second node we should do extra work. We
> - * should mention that it is rare and is a corner case from
> - * the other hand it has to be covered.
> + * VA resides in node 1 whereas it spans 1, 2 an 0. If passed
> + * addr is within 2 or 0 nodes we should do extra work.
> */
> i = j = addr_to_node_id(addr);
> do {
> @@ -2289,6 +2305,9 @@ static struct vmap_area *find_unlink_vmap_area(unsigned long addr)
> struct vmap_area *va;
> int i, j;
>
> + /*
> + * Check the comment in the find_vmap_area() about the loop.
> + */
> i = j = addr_to_node_id(addr);
> do {
> vn = &vmap_nodes[i];
> @@ -4882,7 +4901,20 @@ static void vmap_init_nodes(void)
> int i, n;
>
> #if BITS_PER_LONG == 64
> - /* A high threshold of max nodes is fixed and bound to 128. */
> + /*
> + * A high threshold of max nodes is fixed and bound to 128,
> + * thus a scale factor is 1 for systems where number of cores
> + * are less or equal to specified threshold.
> + *
> + * As for NUMA-aware notes. For bigger systems, for example
> + * NUMA with multi-sockets, where we can end-up with thousands
> + * of cores in total, a "sub-numa-clustering" should be added.
> + *
> + * In this case a NUMA domain is considered as a single entity
> + * with dedicated sub-nodes in it which describe one group or
> + * set of cores. Therefore a per-domain purging is supposed to
> + * be added as well as a per-domain balancing.
> + */
> n = clamp_t(unsigned int, num_possible_cpus(), 1, 128);
>
> if (n > 1) {
> --
> 2.39.2
>
Looks good to me (sorry for delay, busy with many things in life :)! Feel free to add:
Reviewed-by: Lorenzo Stoakes <lstoakes@xxxxxxxxx>
