| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Wed, 21 Dec 2022 18:44:52 +0100
From: "Uladzislau Rezki (Sony)" <urezki@...il.com>
To: Andrew Morton <akpm@...ux-foundation.org>
Cc: linux-mm@...ck.org, LKML <linux-kernel@...r.kernel.org>,
Baoquan He <bhe@...hat.com>,
Lorenzo Stoakes <lstoakes@...il.com>,
Christoph Hellwig <hch@...radead.org>,
Matthew Wilcox <willy@...radead.org>,
Nicholas Piggin <npiggin@...il.com>,
Uladzislau Rezki <urezki@...il.com>,
Oleksiy Avramchenko <oleksiy.avramchenko@...y.com>,
Roman Gushchin <roman.gushchin@...ux.dev>
Subject: [PATCH v2 1/3] mm: vmalloc: Avoid of calling __find_vmap_area() twise in __vunmap()
Currently __vunmap() path calls __find_vmap_area() two times. One on
entry to check that area exists, second time inside remove_vm_area()
function that also performs a new search of VA.
In order to improvie it from a performance point of view we split
remove_vm_area() into two new parts:
- find_unlink_vmap_area() that does a search and unlink from tree;
- __remove_vm_area() that does a removing but without searching.
In this case there is no any functional change for remove_vm_area()
whereas vm_remove_mappings(), where a second search happens, switches
to the __remove_vm_area() variant where already detached VA is passed
as a parameter, so there is no need to find it again.
Performance wise, i use test_vmalloc.sh with 32 threads doing alloc
free on a 64-CPUs-x86_64-box:
perf without this patch:
- 31.41% 0.50% vmalloc_test/10 [kernel.vmlinux] [k] __vunmap
- 30.92% __vunmap
- 17.67% _raw_spin_lock
native_queued_spin_lock_slowpath
- 12.33% remove_vm_area
- 11.79% free_vmap_area_noflush
- 11.18% _raw_spin_lock
native_queued_spin_lock_slowpath
0.76% free_unref_page
perf with this patch:
- 11.35% 0.13% vmalloc_test/14 [kernel.vmlinux] [k] __vunmap
- 11.23% __vunmap
- 8.28% find_unlink_vmap_area
- 7.95% _raw_spin_lock
7.44% native_queued_spin_lock_slowpath
- 1.93% free_vmap_area_noflush
- 0.56% _raw_spin_lock
0.53% native_queued_spin_lock_slowpath
0.60% __vunmap_range_noflush
__vunmap() consumes around ~20% less CPU cycles on this test.
Reported-by: Roman Gushchin <roman.gushchin@...ux.dev>
Signed-off-by: Uladzislau Rezki (Sony) <urezki@...il.com>
---
mm/vmalloc.c | 66 +++++++++++++++++++++++++++++++++-------------------
1 file changed, 42 insertions(+), 24 deletions(-)
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index 9e30f0b39203..28030d2441f1 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -1825,9 +1825,11 @@ static void free_vmap_area_noflush(struct vmap_area *va)
unsigned long va_start = va->va_start;
unsigned long nr_lazy;
- spin_lock(&vmap_area_lock);
- unlink_va(va, &vmap_area_root);
- spin_unlock(&vmap_area_lock);
+ if (!list_empty(&va->list)) {
+ spin_lock(&vmap_area_lock);
+ unlink_va(va, &vmap_area_root);
+ spin_unlock(&vmap_area_lock);
+ }
nr_lazy = atomic_long_add_return((va->va_end - va->va_start) >>
PAGE_SHIFT, &vmap_lazy_nr);
@@ -1871,6 +1873,19 @@ struct vmap_area *find_vmap_area(unsigned long addr)
return va;
}
+static struct vmap_area *find_unlink_vmap_area(unsigned long addr)
+{
+ struct vmap_area *va;
+
+ spin_lock(&vmap_area_lock);
+ va = __find_vmap_area(addr, &vmap_area_root);
+ if (va)
+ unlink_va(va, &vmap_area_root);
+ spin_unlock(&vmap_area_lock);
+
+ return va;
+}
+
/*** Per cpu kva allocator ***/
/*
@@ -2591,6 +2606,20 @@ struct vm_struct *find_vm_area(const void *addr)
return va->vm;
}
+static struct vm_struct *__remove_vm_area(struct vmap_area *va)
+{
+ struct vm_struct *vm;
+
+ if (!va || !va->vm)
+ return NULL;
+
+ vm = va->vm;
+ kasan_free_module_shadow(vm);
+ free_unmap_vmap_area(va);
+
+ return vm;
+}
+
/**
* remove_vm_area - find and remove a continuous kernel virtual area
* @addr: base address
@@ -2607,22 +2636,8 @@ struct vm_struct *remove_vm_area(const void *addr)
might_sleep();
- spin_lock(&vmap_area_lock);
- va = __find_vmap_area((unsigned long)addr, &vmap_area_root);
- if (va && va->vm) {
- struct vm_struct *vm = va->vm;
-
- va->vm = NULL;
- spin_unlock(&vmap_area_lock);
-
- kasan_free_module_shadow(vm);
- free_unmap_vmap_area(va);
-
- return vm;
- }
-
- spin_unlock(&vmap_area_lock);
- return NULL;
+ va = find_unlink_vmap_area((unsigned long) addr);
+ return __remove_vm_area(va);
}
static inline void set_area_direct_map(const struct vm_struct *area,
@@ -2637,15 +2652,16 @@ static inline void set_area_direct_map(const struct vm_struct *area,
}
/* Handle removing and resetting vm mappings related to the vm_struct. */
-static void vm_remove_mappings(struct vm_struct *area, int deallocate_pages)
+static void vm_remove_mappings(struct vmap_area *va, int deallocate_pages)
{
+ struct vm_struct *area = va->vm;
unsigned long start = ULONG_MAX, end = 0;
unsigned int page_order = vm_area_page_order(area);
int flush_reset = area->flags & VM_FLUSH_RESET_PERMS;
int flush_dmap = 0;
int i;
- remove_vm_area(area->addr);
+ __remove_vm_area(va);
/* If this is not VM_FLUSH_RESET_PERMS memory, no need for the below. */
if (!flush_reset)
@@ -2690,6 +2706,7 @@ static void vm_remove_mappings(struct vm_struct *area, int deallocate_pages)
static void __vunmap(const void *addr, int deallocate_pages)
{
struct vm_struct *area;
+ struct vmap_area *va;
if (!addr)
return;
@@ -2698,19 +2715,20 @@ static void __vunmap(const void *addr, int deallocate_pages)
addr))
return;
- area = find_vm_area(addr);
- if (unlikely(!area)) {
+ va = find_unlink_vmap_area((unsigned long)addr);
+ if (unlikely(!va)) {
WARN(1, KERN_ERR "Trying to vfree() nonexistent vm area (%p)\n",
addr);
return;
}
+ area = va->vm;
debug_check_no_locks_freed(area->addr, get_vm_area_size(area));
debug_check_no_obj_freed(area->addr, get_vm_area_size(area));
kasan_poison_vmalloc(area->addr, get_vm_area_size(area));
- vm_remove_mappings(area, deallocate_pages);
+ vm_remove_mappings(va, deallocate_pages);
if (deallocate_pages) {
int i;
--
2.30.2
Powered by blists - more mailing lists