On Wed, Jan 17, 2024 at 09:23:19AM +0000, Chengming Zhou wrote: > Each swapfile has one rb-tree to search the mapping of swp_entry_t to > zswap_entry, that use a spinlock to protect, which can cause heavy lock > contention if multiple tasks zswap_store/load concurrently. > > Optimize the scalability problem by splitting the zswap rb-tree into > multiple rb-trees, each corresponds to SWAP_ADDRESS_SPACE_PAGES (64M), > just like we did in the swap cache address_space splitting. > > Although this method can't solve the spinlock contention completely, it > can mitigate much of that contention. Below is the results of kernel build > in tmpfs with zswap shrinker enabled: > > linux-next zswap-lock-optimize > real 1m9.181s 1m3.820s > user 17m44.036s 17m40.100s > sys 7m37.297s 4m54.622s > > So there are clearly improvements. > > Signed-off-by: Chengming Zhou <zhouchengming@xxxxxxxxxxxxx> Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx> One minor nit: > @@ -265,6 +266,10 @@ static bool zswap_has_pool; > * helpers and fwd declarations > **********************************/ > > +#define swap_zswap_tree(entry) \ > + (&zswap_trees[swp_type(entry)][swp_offset(entry) \ > + >> SWAP_ADDRESS_SPACE_SHIFT]) Make this a static inline function instead?
