Patchv4 was updated based on Hailong and Uladzislau's comments, where
vbq is obtained from vb->cpu, thus avoiding disabling preemption.
Furthermore, Baoquan's suggestion was not adopted because it made vbq
accesses completely interleaved across all CPUs, which defeats the
goal of per_cpu.
On Fri, Jun 7, 2024 at 10:31 AM zhaoyang.huang
<zhaoyang.huang@xxxxxxxxxx> wrote:
>
> From: Zhaoyang Huang <zhaoyang.huang@xxxxxxxxxx>
>
> vmalloc area runs out in our ARM64 system during an erofs test as
> vm_map_ram failed[1]. By following the debug log, we find that
> vm_map_ram()->vb_alloc() will allocate new vb->va which corresponding
> to 4MB vmalloc area as list_for_each_entry_rcu returns immediately
> when vbq->free->next points to vbq->free. That is to say, 65536 times
> of page fault after the list's broken will run out of the whole
> vmalloc area. This should be introduced by one vbq->free->next point to
> vbq->free which makes list_for_each_entry_rcu can not iterate the list
> and find the BUG.
>
> [1]
> PID: 1 TASK: ffffff80802b4e00 CPU: 6 COMMAND: "init"
> #0 [ffffffc08006afe0] __switch_to at ffffffc08111d5cc
> #1 [ffffffc08006b040] __schedule at ffffffc08111dde0
> #2 [ffffffc08006b0a0] schedule at ffffffc08111e294
> #3 [ffffffc08006b0d0] schedule_preempt_disabled at ffffffc08111e3f0
> #4 [ffffffc08006b140] __mutex_lock at ffffffc08112068c
> #5 [ffffffc08006b180] __mutex_lock_slowpath at ffffffc08111f8f8
> #6 [ffffffc08006b1a0] mutex_lock at ffffffc08111f834
> #7 [ffffffc08006b1d0] reclaim_and_purge_vmap_areas at ffffffc0803ebc3c
> #8 [ffffffc08006b290] alloc_vmap_area at ffffffc0803e83fc
> #9 [ffffffc08006b300] vm_map_ram at ffffffc0803e78c0
>
> Fixes: fc1e0d980037 ("mm/vmalloc: prevent stale TLBs in fully utilized blocks")
>
> For detailed reason of broken list, please refer to below URL
> https://lore.kernel.org/all/20240531024820.5507-1-hailong.liu@xxxxxxxx/
>
> Suggested-by: Hailong.Liu <hailong.liu@xxxxxxxx>
> Signed-off-by: Zhaoyang Huang <zhaoyang.huang@xxxxxxxxxx>
> ---
> v2: introduce cpu in vmap_block to record the right CPU number
> v3: use get_cpu/put_cpu to prevent schedule between core
> v4: replace get_cpu/put_cpu by another API to avoid disabling preemption
> ---
> ---
> mm/vmalloc.c | 21 +++++++++++++++------
> 1 file changed, 15 insertions(+), 6 deletions(-)
>
> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> index 22aa63f4ef63..89eb034f4ac6 100644
> --- a/mm/vmalloc.c
> +++ b/mm/vmalloc.c
> @@ -2458,6 +2458,7 @@ struct vmap_block {
> struct list_head free_list;
> struct rcu_head rcu_head;
> struct list_head purge;
> + unsigned int cpu;
> };
>
> /* Queue of free and dirty vmap blocks, for allocation and flushing purposes */
> @@ -2585,8 +2586,15 @@ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask)
> free_vmap_area(va);
> return ERR_PTR(err);
> }
> -
> - vbq = raw_cpu_ptr(&vmap_block_queue);
> + /*
> + * list_add_tail_rcu could happened in another core
> + * rather than vb->cpu due to task migration, which
> + * is safe as list_add_tail_rcu will ensure the list's
> + * integrity together with list_for_each_rcu from read
> + * side.
> + */
> + vb->cpu = raw_smp_processor_id();
> + vbq = per_cpu_ptr(&vmap_block_queue, vb->cpu);
> spin_lock(&vbq->lock);
> list_add_tail_rcu(&vb->free_list, &vbq->free);
> spin_unlock(&vbq->lock);
> @@ -2614,9 +2622,10 @@ static void free_vmap_block(struct vmap_block *vb)
> }
>
> static bool purge_fragmented_block(struct vmap_block *vb,
> - struct vmap_block_queue *vbq, struct list_head *purge_list,
> - bool force_purge)
> + struct list_head *purge_list, bool force_purge)
> {
> + struct vmap_block_queue *vbq = &per_cpu(vmap_block_queue, vb->cpu);
> +
> if (vb->free + vb->dirty != VMAP_BBMAP_BITS ||
> vb->dirty == VMAP_BBMAP_BITS)
> return false;
> @@ -2664,7 +2673,7 @@ static void purge_fragmented_blocks(int cpu)
> continue;
>
> spin_lock(&vb->lock);
> - purge_fragmented_block(vb, vbq, &purge, true);
> + purge_fragmented_block(vb, &purge, true);
> spin_unlock(&vb->lock);
> }
> rcu_read_unlock();
> @@ -2801,7 +2810,7 @@ static void _vm_unmap_aliases(unsigned long start, unsigned long end, int flush)
> * not purgeable, check whether there is dirty
> * space to be flushed.
> */
> - if (!purge_fragmented_block(vb, vbq, &purge_list, false) &&
> + if (!purge_fragmented_block(vb, &purge_list, false) &&
> vb->dirty_max && vb->dirty != VMAP_BBMAP_BITS) {
> unsigned long va_start = vb->va->va_start;
> unsigned long s, e;
> --
> 2.25.1
>
