mm-vmalloc-remove-a-global-vmap_blocks-xarray.patch removed from -mm tree
The quilt patch titled
Subject: mm: vmalloc: remove a global vmap_blocks xarray
has been removed from the -mm tree. Its filename was
mm-vmalloc-remove-a-global-vmap_blocks-xarray.patch
This patch was dropped because it was merged into the mm-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
------------------------------------------------------
From: "Uladzislau Rezki (Sony)" <[email protected]>
Subject: mm: vmalloc: remove a global vmap_blocks xarray
Date: Thu, 30 Mar 2023 21:06:38 +0200
A global vmap_blocks-xarray array can be contented under heavy usage of the vm_map_ram()/vm_unmap_ram() APIs. The lock_stat shows that a "vmap_blocks.xa_lock" lock is a second in a top-list when it comes to contentions:
<snip>
----------------------------------------
class name con-bounces contentions ...
----------------------------------------vmap_area_lock: 2554079 2554276 ...
--------------
vmap_area_lock 1297948 [<00000000dd41cbaa>] alloc_vmap_area+0x1c7/0x910 vmap_area_lock 1256330 [<000000009d927bf3>] free_vmap_block+0x4a/0xe0 vmap_area_lock 1 [<00000000c95c05a7>] find_vm_area+0x16/0x70 -------------- vmap_area_lock 1738590 [<00000000dd41cbaa>] alloc_vmap_area+0x1c7/0x910 vmap_area_lock 815688 [<000000009d927bf3>] free_vmap_block+0x4a/0xe0 vmap_area_lock 1 [<00000000c1d619d7>] __get_vm_area_node+0xd2/0x170
vmap_blocks.xa_lock: 862689 862698 ...
-------------------
vmap_blocks.xa_lock 378418 [<00000000625a5626>] vm_map_ram+0x359/0x4a0 vmap_blocks.xa_lock 484280 [<00000000caa2ef03>] xa_erase+0xe/0x30 ------------------- vmap_blocks.xa_lock 576226 [<00000000caa2ef03>] xa_erase+0xe/0x30 vmap_blocks.xa_lock 286472 [<00000000625a5626>] vm_map_ram+0x359/0x4a0
...
<snip>
that is a result of running vm_map_ram()/vm_unmap_ram() in a loop. The test creates 64(on 64 CPUs system) threads and each one maps/unmaps 1 page.
After this change the "xa_lock" can be considered as a noise in the same test condition:
<snip>
...
&xa->xa_lock#1: 10333 10394 ...
--------------
&xa->xa_lock#1 5349 [<00000000bbbc9751>] xa_erase+0xe/0x30 &xa->xa_lock#1 5045 [<0000000018def45d>] vm_map_ram+0x3a4/0x4f0 -------------- &xa->xa_lock#1 7326 [<0000000018def45d>] vm_map_ram+0x3a4/0x4f0 &xa->xa_lock#1 3068 [<00000000bbbc9751>] xa_erase+0xe/0x30
...
<snip>
Running the test_vmalloc.sh run_test_mask=1024 nr_threads=64 nr_pages=5 shows around ~8 percent of throughput improvement of vm_map_ram() and vm_unmap_ram() APIs.
This patch does not fix vmap_area_lock/free_vmap_area_lock and purge_vmap_area_lock bottle-necks, it is rather a separate rework.
---
mm/vmalloc.c | 71 +++++++++++++++++++++++++++++++++++++++++++------ 1 file changed, 63 insertions(+), 8 deletions(-)
mm/vmalloc.c renamed from mm/vmalloc.c~mm-vmalloc-remove-a-global-vmap_blocks-xarray +63 -8
@@ -1915,6 +1915,13 @@ static struct vmap_area *find_unlink_vmastruct vmap_block_queue {spinlock_t lock;struct list_head free; + +/* +* An xarray requires an extra memory dynamically to +* be allocated. If it is an issue, we can use rb-tree +* instead. +*/ +struct xarray vmap_blocks; };struct vmap_block {
@@ -1932,11 +1939,48 @@ struct vmap_block { static DEFINE_PER_CPU(struct vmap_block_queue, vmap_block_queue);/* -* XArray of vmap blocks, indexed by address, to quickly find a vmap block -* in the free path. Could get rid of this if we change the API to return a -* "cookie" from alloc, to be passed to free. But no big deal yet. +* In order to fast access to any "vmap_block" associated with a +* specific address, we use a hash. +* +* A per-cpu vmap_block_queue is used in both ways, to serialize +* an access to free block chains among CPUs(alloc path) and it +* also acts as a vmap_block hash(alloc/free paths). It means we +* overload it, since we already have the per-cpu array which is +* used as a hash table. When used as a hash a 'cpu' passed to +* per_cpu() is not actually a CPU but rather a hash index. +* +* A hash function is addr_to_vb_xarray() which hashes any address +* to a specific index(in a hash) it belongs to. This then uses a +* per_cpu() macro to access an array with generated index. +* +* An example: +* +* CPU_1 CPU_2 CPU_0 +* | | | +* V V V +* 0 10 20 30 40 50 60 +* |------|------|------|------|------|------|...<vmap address space> +* CPU0 CPU1 CPU2 CPU0 CPU1 CPU2 +* +* - CPU_1 invokes vm_unmap_ram(6), 6 belongs to CPU0 zone, thus +* it access: CPU0/INDEX0 -> vmap_blocks -> xa_lock; +* +* - CPU_2 invokes vm_unmap_ram(11), 11 belongs to CPU1 zone, thus +* it access: CPU1/INDEX1 -> vmap_blocks -> xa_lock; +* +* - CPU_0 invokes vm_unmap_ram(20), 20 belongs to CPU2 zone, thus +* it access: CPU2/INDEX2 -> vmap_blocks -> xa_lock. +* +* This technique almost always avoids lock contention on insert/remove, +* however xarray spinlocks protect against any contention that remains. */ -static DEFINE_XARRAY (vmap_blocks); +static struct xarray * +addr_to_vb_xarray (unsigned long addr) +{ +int index = (addr /VMAP_BLOCK_SIZE ) %num_possible_cpus (); + +return &per_cpu (vmap_block_queue, index).vmap_blocks; +}/* * We should probably have a fallback mechanism to allocate virtual memory
@@ -1974,6 +2018,7 @@ static void *new_vmap_block(unsigned intstruct vmap_block_queue *vbq;struct vmap_block *vb;struct vmap_area *va; +struct xarray *xa;unsigned long vb_idx;int node, err;void *vaddr;
@@ -2007,8 +2052,9 @@ static void *new_vmap_block(unsigned intbitmap_set (vb->used_map,0 , (1UL << order));INIT_LIST_HEAD (&vb->free_list); + xa =addr_to_vb_xarray (va->va_start); vb_idx =addr_to_vb_idx (va->va_start); - err =xa_insert (&vmap_blocks, vb_idx, vb, gfp_mask); + err =xa_insert (xa, vb_idx, vb, gfp_mask);if (err) {kfree (vb);free_vmap_area (va);
@@ -2026,8 +2072,10 @@ static void *new_vmap_block(unsigned intstatic void free_vmap_block (struct vmap_block *vb) {struct vmap_block *tmp; +struct xarray *xa; - tmp =xa_erase (&vmap_blocks,addr_to_vb_idx (vb->va->va_start)); + xa =addr_to_vb_xarray (vb->va->va_start); + tmp =xa_erase (xa,addr_to_vb_idx (vb->va->va_start));BUG_ON (tmp != vb);spin_lock (&vmap_area_lock);
@@ -2139,6 +2187,7 @@ static void vb_free(unsigned long addr,unsigned long offset;unsigned int order;struct vmap_block *vb; +struct xarray *xa;BUG_ON (offset_in_page (size));BUG_ON (size >PAGE_SIZE *VMAP_MAX_ALLOC );
@@ -2147,7 +2196,10 @@ static void vb_free(unsigned long addr, order =get_order (size); offset = (addr & (VMAP_BLOCK_SIZE -1 )) >>PAGE_SHIFT ; - vb =xa_load (&vmap_blocks,addr_to_vb_idx (addr)); + + xa =addr_to_vb_xarray (addr); + vb =xa_load (xa,addr_to_vb_idx (addr)); +spin_lock (&vb->lock);bitmap_clear (vb->used_map, offset, (1UL << order));spin_unlock (&vb->lock);
@@ -3525,6 +3577,7 @@ static size_t vmap_ram_vread_iter(struct {char *start;struct vmap_block *vb; +struct xarray *xa;unsigned long offset;unsigned int rs, re;size_t remains, n;
@@ -3543,7 +3596,8 @@ static size_t vmap_ram_vread_iter(struct* Area is split into regions and tracked with vmap_block, read out * each region and zero fill the hole between regions. */ - vb =xa_load (&vmap_blocks,addr_to_vb_idx ((unsigned long )addr)); + xa =addr_to_vb_xarray ((unsigned long ) addr); + vb =xa_load (xa,addr_to_vb_idx ((unsigned long )addr));if (!vb) goto finished_zero;
@@ -4337,6 +4391,7 @@ void __init vmalloc_init(void) p = &per_cpu (vfree_deferred, i);init_llist_head (&p->list);INIT_WORK (&p->wq, delayed_vfree_work); +xa_init (&vbq->vmap_blocks); }/* Import existing vmlist entries. */
_
Patches currently in -mm which might be from [email protected] are
lib-test_vmallocc-add-vm_map_ram-vm_unmap_ram-test-case.patch mm-vmalloc-rename-addr_to_vb_xarray-function.patch