Re: [PATCH V3 02/12] perf mmap: factor out function to find ringbuffer position

[Jiri Olsa]

On Thu, Dec 21, 2017 at 10:08:44AM -0800, kan.liang@xxxxxxxxx wrote:

SNIP

> +/*
> + * Report the start and end of the available data in ringbuffer
> + */
> +int perf_mmap__read_init(struct perf_mmap *map, bool overwrite,
> +			 u64 *start, u64 *end)
>  {
> -	u64 head = perf_mmap__read_head(md);
> -	u64 old = md->prev;
> -	u64 end = head, start = old;
> -	unsigned char *data = md->base + page_size;
> +	u64 head = perf_mmap__read_head(map);
> +	u64 old = map->prev;
> +	unsigned char *data = map->base + page_size;
>  	unsigned long size;
> -	void *buf;
> -	int rc = 0;
>  
> -	start = overwrite ? head : old;
> -	end = overwrite ? old : head;
> +	/*
> +	 * Check if event was unmapped due to a POLLHUP/POLLERR.
> +	 */
> +	if (!refcount_read(&map->refcnt))
> +		return -EINVAL;
>  
> -	if (start == end)
> -		return 0;
> +	*start = overwrite ? head : old;
> +	*end = overwrite ? old : head;
>  
> -	size = end - start;
> -	if (size > (unsigned long)(md->mask) + 1) {
> +	if (*start == *end)
> +		return -EAGAIN;
> +
> +	size = *end - *start;
> +	if (size > (unsigned long)(map->mask) + 1) {
>  		if (!overwrite) {
>  			WARN_ONCE(1, "failed to keep up with mmap data. (warn only once)\n");
>  

I know you did not change this, but is this leg even possible
in !overwrite mode? I think kernel will throw away the data,
keep the head and wait for tail to be read by user..

jirka

> -			md->prev = head;
> -			perf_mmap__consume(md, overwrite);
> -			return 0;
> +			map->prev = head;
> +			perf_mmap__consume(map, overwrite);
> +			return -EAGAIN;

SNIP