Hi Kairui,
On 02/15/25 at 01:57am, Kairui Song wrote:
......snip....
> -int get_swap_pages(int n_goal, swp_entry_t swp_entries[], int entry_order)
> +swp_entry_t folio_alloc_swap(struct folio *folio)
> {
> - int order = swap_entry_order(entry_order);
> - unsigned long size = 1 << order;
> + unsigned int order = folio_order(folio);
> + unsigned int size = 1 << order;
> struct swap_info_struct *si, *next;
> - int n_ret = 0;
> + swp_entry_t entry = {};
> + unsigned long offset;
> int node;
>
> + if (order) {
> + /*
> + * Should not even be attempting large allocations when huge
> + * page swap is disabled. Warn and fail the allocation.
> + */
> + if (!IS_ENABLED(CONFIG_THP_SWAP) || size > SWAPFILE_CLUSTER) {
> + VM_WARN_ON_ONCE(1);
> + return entry;
> + }
> + }
> +
> /* Fast path using percpu cluster */
> local_lock(&percpu_swap_cluster.lock);
> - n_ret = swap_alloc_fast(swp_entries,
> - SWAP_HAS_CACHE,
> - order, n_goal);
> - if (n_ret == n_goal)
> - goto out;
> + if (swap_alloc_fast(&entry, SWAP_HAS_CACHE, order))
> + goto out_alloced;
>
> - n_goal = min_t(int, n_goal - n_ret, SWAP_BATCH);
> /* Rotate the device and switch to a new cluster */
> spin_lock(&swap_avail_lock);
> start_over:
> @@ -1268,11 +1236,14 @@ int get_swap_pages(int n_goal, swp_entry_t swp_entries[], int entry_order)
> plist_requeue(&si->avail_lists[node], &swap_avail_heads[node]);
> spin_unlock(&swap_avail_lock);
> if (get_swap_device_info(si)) {
> - n_ret += scan_swap_map_slots(si, SWAP_HAS_CACHE, n_goal,
> - swp_entries + n_ret, order);
> + offset = cluster_alloc_swap_entry(si, order, SWAP_HAS_CACHE);
> put_swap_device(si);
> - if (n_ret || size > 1)
> - goto out;
> + if (offset) {
> + entry = swp_entry(si->type, offset);
> + goto out_alloced;
> + }
> + if (order)
> + goto out_failed;
This is not related to this patch, do you know why non order-0 case
can't start over on different devices?
> }
>
> spin_lock(&swap_avail_lock);
> @@ -1291,10 +1262,20 @@ int get_swap_pages(int n_goal, swp_entry_t swp_entries[], int entry_order)
> goto start_over;
> }
> spin_unlock(&swap_avail_lock);
> -out:
> +out_failed:
> + local_unlock(&percpu_swap_cluster.lock);
> + return entry;
> +
> +out_alloced:
> local_unlock(&percpu_swap_cluster.lock);
> - atomic_long_sub(n_ret * size, &nr_swap_pages);
> - return n_ret;
> + if (mem_cgroup_try_charge_swap(folio, entry)) {
> + put_swap_folio(folio, entry);
> + entry.val = 0;
> + } else {
> + atomic_long_sub(size, &nr_swap_pages);
> + }
> +
> + return entry;
> }
>
> static struct swap_info_struct *_swap_info_get(swp_entry_t entry)
......snip....
> @@ -2623,16 +2591,6 @@ static bool __has_usable_swap(void)
> return !plist_head_empty(&swap_active_head);
> }
seems the __has_usable_swap() function need be moved into the ifdeffery
scope where __folio_throttle_swaprate() is located to fix the lkp
warning.
>
> -bool has_usable_swap(void)
> -{
> - bool ret;
> -
> - spin_lock(&swap_lock);
> - ret = __has_usable_swap();
> - spin_unlock(&swap_lock);
> - return ret;
> -}
> -
> /*
> * Called after clearing SWP_WRITEOK, ensures cluster_alloc_range
> * see the updated flags, so there will be no more allocations.
Other than the test robot reported warning, this patch looks good to me.
Thanks.
