> On Aug 23, 2023, at 8:42 PM, Qi Zheng <zhengqi.arch@xxxxxxxxxxxxx> wrote:
>
> In preparation for implementing lockless slab shrink, use new APIs to
> dynamically allocate the vmw-balloon shrinker, so that it can be freed
> asynchronously via RCU. Then it doesn't need to wait for RCU read-side
> critical section when releasing the struct vmballoon.
>
> And we can simply exit vmballoon_init() when registering the shrinker
> fails. So the shrinker_registered indication is redundant, just remove it.
...
Ugh. We should have already moved to OOM notifier instead...
> static void vmballoon_unregister_shrinker(struct vmballoon *b)
> {
> - if (b->shrinker_registered)
> - unregister_shrinker(&b->shrinker);
> - b->shrinker_registered = false;
> + shrinker_free(b->shrinker);
> }
If the patch goes through another iteration, please add:
b->shrinker = NULL;
Not that this is a real issue, but I prefer it so in order to more easily
identify UAF if the function is called elsewhere.
Otherwise, LGTM. Thanks.
