Re: [PATCH bpf-next v6 1/5] bpf: Introduce task_file open-coded iterator kfuncs

[Yonghong Song <yonghong.song@linux.dev>]

On 12/17/24 3:37 PM, Juntong Deng wrote:
> This patch adds the open-coded iterator style process file iterator
> kfuncs bpf_iter_task_file_{new,next,destroy} that iterates over all
> files opened by the specified process.
>
> bpf_iter_task_file_next returns a pointer to bpf_iter_task_file_item,
> which currently contains *task, *file, fd. This is an extensible
> structure that enables compatibility with different versions
> through CO-RE.
>
> The reference to struct file acquired by the previous
> bpf_iter_task_file_next() is released in the next
> bpf_iter_task_file_next(), and the last reference is released in the
> last bpf_iter_task_file_next() that returns NULL.
>
> In the bpf_iter_task_file_destroy(), if the iterator does not iterate to
> the end, then the last struct file reference is released at this time.
>
> Signed-off-by: Juntong Deng <juntong.deng@outlook.com>
> ---
>   kernel/bpf/helpers.c   |  3 ++
>   kernel/bpf/task_iter.c | 91 ++++++++++++++++++++++++++++++++++++++++++
>   2 files changed, 94 insertions(+)
>
> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> index cd5f9884d85b..61a652bea0ba 100644
> --- a/kernel/bpf/helpers.c
> +++ b/kernel/bpf/helpers.c
> @@ -3147,6 +3147,9 @@ BTF_ID_FLAGS(func, bpf_iter_css_destroy, KF_ITER_DESTROY)
>   BTF_ID_FLAGS(func, bpf_iter_task_new, KF_ITER_NEW | KF_TRUSTED_ARGS | KF_RCU_PROTECTED)
>   BTF_ID_FLAGS(func, bpf_iter_task_next, KF_ITER_NEXT | KF_RET_NULL)
>   BTF_ID_FLAGS(func, bpf_iter_task_destroy, KF_ITER_DESTROY)
> +BTF_ID_FLAGS(func, bpf_iter_task_file_new, KF_ITER_NEW | KF_TRUSTED_ARGS)
> +BTF_ID_FLAGS(func, bpf_iter_task_file_next, KF_ITER_NEXT | KF_RET_NULL)
> +BTF_ID_FLAGS(func, bpf_iter_task_file_destroy, KF_ITER_DESTROY)
>   BTF_ID_FLAGS(func, bpf_dynptr_adjust)
>   BTF_ID_FLAGS(func, bpf_dynptr_is_null)
>   BTF_ID_FLAGS(func, bpf_dynptr_is_rdonly)
> diff --git a/kernel/bpf/task_iter.c b/kernel/bpf/task_iter.c
> index 98d9b4c0daff..149a95762f68 100644
> --- a/kernel/bpf/task_iter.c
> +++ b/kernel/bpf/task_iter.c
> @@ -1027,6 +1027,97 @@ __bpf_kfunc void bpf_iter_task_destroy(struct bpf_iter_task *it)
>   {
>   }
>   
> +struct bpf_iter_task_file_item {
> +	struct task_struct *task;
> +	struct file *file;
> +	unsigned int fd;
> +} __aligned(8);

We probably do not __aligned(8) here as alignment has been
guaranteed in struct bpf_iter_task_file_kern.

> +
> +struct bpf_iter_task_file {
> +	__u64 __opaque[4];
> +} __aligned(8);
> +
> +struct bpf_iter_task_file_kern {
> +	struct bpf_iter_task_file_item item;
> +	unsigned int next_fd;
> +} __aligned(8);
> +
> +/**
> + * bpf_iter_task_file_new() - Initialize a new task file iterator for a task,
> + * used to iterate over all files opened by a specified task
> + *
> + * @it: the new bpf_iter_task_file to be created
> + * @task: a pointer pointing to a task to be iterated over
> + */
> +__bpf_kfunc int bpf_iter_task_file_new(struct bpf_iter_task_file *it, struct task_struct *task)
> +{
> +	struct bpf_iter_task_file_kern *kit = (void *)it;
> +	struct bpf_iter_task_file_item *item = &kit->item;
> +
> +	BUILD_BUG_ON(sizeof(struct bpf_iter_task_file_kern) > sizeof(struct bpf_iter_task_file));
> +	BUILD_BUG_ON(__alignof__(struct bpf_iter_task_file_kern) !=
> +		     __alignof__(struct bpf_iter_task_file));
> +
> +	item->task = get_task_struct(task);
> +	item->file = NULL;
> +	item->fd = 0;
> +	kit->next_fd = 0;
> +
> +	return 0;
> +}
> +
> +/**
> + * bpf_iter_task_file_next() - Get the next file in bpf_iter_task_file
> + *
> + * bpf_iter_task_file_next acquires a reference to the struct file.
> + *
> + * The reference to struct file acquired by the previous
> + * bpf_iter_task_file_next() is released in the next bpf_iter_task_file_next(),
> + * and the last reference is released in the last bpf_iter_task_file_next()
> + * that returns NULL.
> + *
> + * @it: the bpf_iter_task_file to be checked
> + *
> + * @returns a pointer to bpf_iter_task_file_item
> + */
> +__bpf_kfunc struct bpf_iter_task_file_item *bpf_iter_task_file_next(struct bpf_iter_task_file *it)
> +{
> +	struct bpf_iter_task_file_kern *kit = (void *)it;
> +	struct bpf_iter_task_file_item *item = &kit->item;
> +
> +	if (item->file)
> +		fput(item->file);
> +
> +	item->file = fget_task_next(item->task, &kit->next_fd);
> +	item->fd = kit->next_fd;
> +
> +	kit->next_fd++;
> +
> +	if (!item->file)
> +		return NULL;

Maybe move the above if statement right after
	iterm->file = fget_task_next(item->task, &kit->next_fd);
to make code more coherent?

> +
> +	return item;
> +}
> +
> +/**
> + * bpf_iter_task_file_destroy() - Destroy a bpf_iter_task_file
> + *
> + * If the iterator does not iterate to the end, then the last
> + * struct file reference is released at this time.
> + *
> + * @it: the bpf_iter_task_file to be destroyed
> + */
> +__bpf_kfunc void bpf_iter_task_file_destroy(struct bpf_iter_task_file *it)
> +{
> +	struct bpf_iter_task_file_kern *kit = (void *)it;
> +	struct bpf_iter_task_file_item *item = &kit->item;
> +
> +	if (item->file)
> +		fput(item->file);
> +
> +	put_task_struct(item->task);
> +}
> +
>   __bpf_kfunc_end_defs();
>   
>   DEFINE_PER_CPU(struct mmap_unlock_irq_work, mmap_unlock_work);