Re: [PATCH] mm: fix racing of vb->va when kasan enabled
From: Uladzislau Rezki
Date: Fri Jun 24 2022 - 06:27:58 EST
> On Wed, Jun 22, 2022 at 11:15 AM Zhaoyang Huang <huangzhaoyang@xxxxxxxxx> wrote:
> >
> > On Tue, Jun 21, 2022 at 10:29 PM Uladzislau Rezki <urezki@xxxxxxxxx> wrote:
> > >
> > > > On Tue, Jun 21, 2022 at 5:27 PM Uladzislau Rezki <urezki@xxxxxxxxx> wrote:
> > > > >
> > > > > > On Mon, Jun 20, 2022 at 6:44 PM Uladzislau Rezki <urezki@xxxxxxxxx> wrote:
> > > > > > >
> > > > > > > > > >
> > > > > > > > > Is it easy to reproduce? If so could you please describe the steps? As i see
> > > > > > > > > the freeing of the "vb" is RCU safe whereas vb->va is not. But from the first
> > > > > > > > > glance i do not see how it can accessed twice. Hm..
> > > > > > > > It was raised from a monkey test on A13_k515 system and got 1/20 pcs
> > > > > > > > failed. IMO, vb->va which out of vmap_purge_lock protection could race
> > > > > > > > with a concurrent ra freeing within __purge_vmap_area_lazy.
> > > > > > > >
> > > > > > > Do you have exact steps how you run "monkey" test?
> > > > > > There are about 30+ kos inserted during startup which could be a
> > > > > > specific criteria for reproduction. Do you have doubts about the test
> > > > > > result or the solution?
> > > > > > >
> > > > > I do not have any doubt about your test results, so if you can trigger it
> > > > > then there is an issue at least on the 5.4.161-android12 kernel.
> > > > >
> > > > > 1. With your fix we get expanded mutex range, thus the worst case of vmalloc
> > > > > allocation can be increased when it fails and repeat. Because it also invokes
> > > > > the purge_vmap_area_lazy() that access the same mutex.
> > > > I am not sure I get your point. _vm_unmap_aliases calls
> > > > _purge_vmap_area_lazy instead of purge_vmap_area_lazy. Do you have any
> > > > other solutions? I really don't think my patch is the best way as I
> > > > don't have a full view of vmalloc mechanism.
> > > >
> > > Yep, but it holds the mutex:
> I still don't get how _purge_vmap_area_lazy hold vmap_purge_lock?
>
The user has to take the mutex if it invokes the __purge_vmap_area_lazy()
function.
> > >
> > > <snip>
> > > mutex_lock(&vmap_purge_lock);
> > > purge_fragmented_blocks_allcpus();
> > > if (!__purge_vmap_area_lazy(start, end) && flush)
> > > flush_tlb_kernel_range(start, end);
> > > mutex_unlock(&vmap_purge_lock);
> > > <snip>
> > >
> > > I do not have a solution yet. I am trying still to figure out how you can
> > > trigger it.
> > >
> > > <snip>
> > > rcu_read_lock();
> > > list_for_each_entry_rcu(vb, &vbq->free, free_list) {
> > > spin_lock(&vb->lock);
> > > if (vb->dirty && vb->dirty != VMAP_BBMAP_BITS) {
> > > unsigned long va_start = vb->va->va_start;
> > > <snip>
> > >
> > > so you say that "vb->va->va_start" can be accessed twice. I do not see
> > > how it can happen. The purge_fragmented_blocks() removes "vb" from the
> > > free_list and set vb->dirty to the VMAP_BBMAP_BITS to prevent purging
> > > it again. It is protected by the spin_lock(&vb->lock):
> > >
> > > <snip>
> > > spin_lock(&vb->lock);
> > > if (vb->free + vb->dirty == VMAP_BBMAP_BITS && vb->dirty != VMAP_BBMAP_BITS) {
> > > vb->free = 0; /* prevent further allocs after releasing lock */
> > > vb->dirty = VMAP_BBMAP_BITS; /* prevent purging it again */
> > > vb->dirty_min = 0;
> > > vb->dirty_max = VMAP_BBMAP_BITS;
> > > <snip>
> > >
> > > so the VMAP_BBMAP_BITS is set under spinlock. The _vm_unmap_aliases() checks it:
> > >
> > > <snip>
> > > list_for_each_entry_rcu(vb, &vbq->free, free_list) {
> > > spin_lock(&vb->lock);
> > > if (vb->dirty && vb->dirty != VMAP_BBMAP_BITS) {
> > > unsigned long va_start = vb->va->va_start;
> > > unsigned long s, e;
> > > <snip>
> > >
> > > if the "vb->dirty != VMAP_BBMAP_BITS". I am missing your point here?
> > Could the racing be like bellowing scenario? vb->va accessed in [2]
> > has been freed in [1]
> >
> > _vm_unmap_aliases
> > _vm_unmap_aliases
> > {
> > {
> > list_for_each_entry_rcu(vb, &vbq->free, free_list) {
> > __purge_vmap_area_lazy
> > spin_lock(&vb->lock);
> > merge_or_add_vmap_area
> > if (vb->dirty) {
> >
> > kmem_cache_free(vmap_area_cachep, va)[1]
> > unsigned long va_start = vb->va->va_start;
> > [2]
>
> reformat the racing graph
> _vm_unmap_aliases
> _vm_unmap_aliases
> {
> {
> list_for_each_entry_rcu(vb, &vbq->free, free_list) {
> __purge_vmap_area_lazy
> spin_lock(&vb->lock);
> merge_or_add_vmap_area
> if (vb->dirty) {
>
> kmem_cache_free(vmap_area_cachep, va)[1]
> unsigned long va_start = vb->va->va_start; [2]
>
> > >
> > > > >
> > > > > 2. You run 5.4.161-android12 kernel what is quite old. Could you please
> > > > > retest with latest kernel? I am asking because on the latest kernel with
> > > > > CONFIG_KASAN i am not able to reproduce it.
> > > > >
> > > > > I do a lot of: vm_map_ram()/vm_unmap_ram()/vmalloc()/vfree() in parallel
> > > > > by 64 kthreads on my 64 CPUs test system.
> > > > The failure generates at 20s from starting up, I think it is a rare timing.
> > > > >
> > > > > Could you please confirm that you can trigger an issue on the latest kernel?
> > > > Sorry, I don't have an available latest kernel for now.
> > > >
> > > Can you do: "gdb ./vmlinux", execute "l *_vm_unmap_aliases+0x164" and provide
> > > output?
> Sorry, I have lost the vmlinux with KASAN enabled and just got some
> instructions from logs.
>
> 0xffffffd010678da8 <_vm_unmap_aliases+0x134>: sub x22, x26, #0x28
> x26 vbq->free
> 0xffffffd010678dac <_vm_unmap_aliases+0x138>: lsr x8, x22, #3
> 0xffffffd010678db0 <_vm_unmap_aliases+0x13c>: ldrb w8, [x8,x24]
> 0xffffffd010678db4 <_vm_unmap_aliases+0x140>: cbz w8,
> 0xffffffd010678dc0 <_vm_unmap_aliases+0x14c>
> 0xffffffd010678db8 <_vm_unmap_aliases+0x144>: mov x0, x22
> 0xffffffd010678dbc <_vm_unmap_aliases+0x148>: bl 0xffffffd0106c9a34
> <__asan_report_load8_noabort>
> 0xffffffd010678dc0 <_vm_unmap_aliases+0x14c>: ldr x22, [x22]
> 0xffffffd010678dc4 <_vm_unmap_aliases+0x150>: lsr x8, x22, #3
> 0xffffffd010678dc8 <_vm_unmap_aliases+0x154>: ldrb w8, [x8,x24]
> 0xffffffd010678dcc <_vm_unmap_aliases+0x158>: cbz w8,
> 0xffffffd010678dd8 <_vm_unmap_aliases+0x164>
> 0xffffffd010678dd0 <_vm_unmap_aliases+0x15c>: mov x0, x22
> 0xffffffd010678dd4 <_vm_unmap_aliases+0x160>: bl 0xffffffd0106c9a34
> <__asan_report_load8_noabort>
>
Could you please test below patch if that fixes an issue on the 5.4
kernel:
<snip>
diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h
index 4e7809408073..d5b07d7239bd 100644
--- a/include/linux/vmalloc.h
+++ b/include/linux/vmalloc.h
@@ -55,6 +55,7 @@ struct vmap_area {
struct rb_node rb_node; /* address sorted rbtree */
struct list_head list; /* address sorted list */
+ struct rcu_head rcu;
/*
* The following three variables can be packed, because
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index a3c70e275f4e..bb8cfdb06ce6 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -337,14 +337,6 @@ static LLIST_HEAD(vmap_purge_list);
static struct rb_root vmap_area_root = RB_ROOT;
static bool vmap_initialized __read_mostly;
-/*
- * This kmem_cache is used for vmap_area objects. Instead of
- * allocating from slab we reuse an object from this cache to
- * make things faster. Especially in "no edge" splitting of
- * free block.
- */
-static struct kmem_cache *vmap_area_cachep;
-
/*
* This linked list is used in pair with free_vmap_area_root.
* It gives O(1) access to prev/next to perform fast coalescing.
@@ -532,7 +524,7 @@ link_va(struct vmap_area *va, struct rb_root *root,
}
/* Address-sort this list */
- list_add(&va->list, head);
+ list_add_rcu(&va->list, head);
}
static __always_inline void
@@ -547,7 +539,7 @@ unlink_va(struct vmap_area *va, struct rb_root *root)
else
rb_erase(&va->rb_node, root);
- list_del(&va->list);
+ list_del_rcu(&va->list);
RB_CLEAR_NODE(&va->rb_node);
}
@@ -721,7 +713,7 @@ merge_or_add_vmap_area(struct vmap_area *va,
augment_tree_propagate_from(sibling);
/* Free vmap_area object. */
- kmem_cache_free(vmap_area_cachep, va);
+ kfree_rcu(va, rcu);
/* Point to the new merged area. */
va = sibling;
@@ -748,7 +740,7 @@ merge_or_add_vmap_area(struct vmap_area *va,
unlink_va(va, root);
/* Free vmap_area object. */
- kmem_cache_free(vmap_area_cachep, va);
+ kfree_rcu(va, rcu);
return;
}
}
@@ -928,7 +920,7 @@ adjust_va_to_fit_type(struct vmap_area *va,
* |---------------|
*/
unlink_va(va, &free_vmap_area_root);
- kmem_cache_free(vmap_area_cachep, va);
+ kfree_rcu(va, rcu);
} else if (type == LE_FIT_TYPE) {
/*
* Split left edge of fit VA.
@@ -969,7 +961,7 @@ adjust_va_to_fit_type(struct vmap_area *va,
* a first allocation (early boot up) when we have "one"
* big free space that has to be split.
*/
- lva = kmem_cache_alloc(vmap_area_cachep, GFP_NOWAIT);
+ lva = kmalloc(sizeof(struct vmap_area), GFP_NOWAIT);
if (!lva)
return -1;
}
@@ -1064,8 +1056,7 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
might_sleep();
- va = kmem_cache_alloc_node(vmap_area_cachep,
- gfp_mask & GFP_RECLAIM_MASK, node);
+ va = kmalloc_node(sizeof(struct vmap_area), gfp_mask & GFP_RECLAIM_MASK, node);
if (unlikely(!va))
return ERR_PTR(-ENOMEM);
@@ -1091,12 +1082,12 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
preempt_disable();
if (!__this_cpu_read(ne_fit_preload_node)) {
preempt_enable();
- pva = kmem_cache_alloc_node(vmap_area_cachep, GFP_KERNEL, node);
+ pva = kmalloc_node(sizeof(struct vmap_area), GFP_KERNEL, node);
preempt_disable();
if (__this_cpu_cmpxchg(ne_fit_preload_node, NULL, pva)) {
if (pva)
- kmem_cache_free(vmap_area_cachep, pva);
+ kfree_rcu(pva, rcu);
}
}
@@ -1145,7 +1136,7 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
pr_warn("vmap allocation for size %lu failed: use vmalloc=<size> to increase size\n",
size);
- kmem_cache_free(vmap_area_cachep, va);
+ kfree_rcu(va, rcu);
return ERR_PTR(-EBUSY);
}
@@ -1870,7 +1861,7 @@ static void vmap_init_free_space(void)
*/
list_for_each_entry(busy, &vmap_area_list, list) {
if (busy->va_start - vmap_start > 0) {
- free = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT);
+ free = kzalloc(sizeof(struct vmap_area), GFP_NOWAIT);
if (!WARN_ON_ONCE(!free)) {
free->va_start = vmap_start;
free->va_end = busy->va_start;
@@ -1885,7 +1876,7 @@ static void vmap_init_free_space(void)
}
if (vmap_end - vmap_start > 0) {
- free = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT);
+ free = kzalloc(sizeof(struct vmap_area), GFP_NOWAIT);
if (!WARN_ON_ONCE(!free)) {
free->va_start = vmap_start;
free->va_end = vmap_end;
@@ -1903,11 +1894,6 @@ void __init vmalloc_init(void)
struct vm_struct *tmp;
int i;
- /*
- * Create the cache for vmap_area objects.
- */
- vmap_area_cachep = KMEM_CACHE(vmap_area, SLAB_PANIC);
-
for_each_possible_cpu(i) {
struct vmap_block_queue *vbq;
struct vfree_deferred *p;
@@ -1922,7 +1908,7 @@ void __init vmalloc_init(void)
/* Import existing vmlist entries. */
for (tmp = vmlist; tmp; tmp = tmp->next) {
- va = kmem_cache_zalloc(vmap_area_cachep, GFP_NOWAIT);
+ va = kzalloc(sizeof(struct vmap_area), GFP_NOWAIT);
if (WARN_ON_ONCE(!va))
continue;
@@ -3256,7 +3242,7 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
goto err_free2;
for (area = 0; area < nr_vms; area++) {
- vas[area] = kmem_cache_zalloc(vmap_area_cachep, GFP_KERNEL);
+ vas[area] = kzalloc(sizeof(struct vmap_area), GFP_KERNEL);
vms[area] = kzalloc(sizeof(struct vm_struct), GFP_KERNEL);
if (!vas[area] || !vms[area])
goto err_free;
@@ -3376,8 +3362,7 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
if (vas[area])
continue;
- vas[area] = kmem_cache_zalloc(
- vmap_area_cachep, GFP_KERNEL);
+ vas[area] = kzalloc(sizeof(struct vmap_area), GFP_KERNEL);
if (!vas[area])
goto err_free;
}
@@ -3388,7 +3373,7 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
err_free:
for (area = 0; area < nr_vms; area++) {
if (vas[area])
- kmem_cache_free(vmap_area_cachep, vas[area]);
+ kfree_rcu(vas[area], rcu);
kfree(vms[area]);
}
<snip>
--
Uladzislau Rezki