Date: Tue, 16 Jan 2018 14:12:37 +0100
From: Jiri Olsa <jolsa@redhat.com>
To: kan.liang@intel.com
Cc: acme@kernel.org, peterz@infradead.org, mingo@redhat.com,
        linux-kernel@vger.kernel.org, wangnan0@huawei.com,
        jolsa@kernel.org, namhyung@kernel.org, ak@linux.intel.com,
        yao.jin@linux.intel.com
Subject: Re: [PATCH V4 02/15] perf mmap: introduce perf_mmap__read_init()
Message-ID: <20180116131237.GG26643@krava>
References: <1516047651-164336-1-git-send-email-kan.liang@intel.com>
 <1516047651-164336-3-git-send-email-kan.liang@intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1516047651-164336-3-git-send-email-kan.liang@intel.com>
User-Agent: Mutt/1.9.1 (2017-09-22)
Sender: linux-kernel-owner@vger.kernel.org

On Mon, Jan 15, 2018 at 12:20:38PM -0800, kan.liang@intel.com wrote:
> From: Kan Liang <kan.liang@intel.com>
> 
> The perf record has specific codes to calculate the ringbuffer position
> for both overwrite and non-overwrite mode. Now, only perf record
> supports both modes. The perf top will support both modes later.
> It is useful to make the specific codes generic.
> 
> Introduce a new interface perf_mmap__read_init() to find ringbuffer
> position. The perf_mmap__read_init() is actually factored out from
> perf_mmap__push().
> There are slight differences.
> - Add a check for map->refcnt
> - Add new return value logic, EAGAIN and EINVAL.

not helpful.. I asked to separate those changes,
so we can clearly see what the refcnt check for
and what's behing that EAGAIN return

please add separate:
  1) patch that adds perf_mmap__read_init into perf_mmap__push
     with no functional change
  2) patch adds adds and explain the refcnt check
  3) patch that adds and explain the EAGAIN return

thanks,
jirka

> 
> Signed-off-by: Kan Liang <kan.liang@intel.com>
> ---
>  tools/perf/util/mmap.c | 43 +++++++++++++++++++++++++++++++++++++++++++
>  tools/perf/util/mmap.h |  2 ++
>  2 files changed, 45 insertions(+)
> 
> diff --git a/tools/perf/util/mmap.c b/tools/perf/util/mmap.c
> index 05076e6..414089f 100644
> --- a/tools/perf/util/mmap.c
> +++ b/tools/perf/util/mmap.c
> @@ -267,6 +267,49 @@ static int overwrite_rb_find_range(void *buf, int mask, u64 head, u64 *start, u6
>  	return -1;
>  }
>  
> +/*
> + * Report the start and end of the available data in ringbuffer
> + */
> +int perf_mmap__read_init(struct perf_mmap *map, bool overwrite,
> +			 u64 *startp, u64 *endp)
> +{
> +	unsigned char *data = map->base + page_size;
> +	u64 head = perf_mmap__read_head(map);
> +	u64 old = map->prev;
> +	unsigned long size;
> +
> +	/*
> +	 * Check if event was unmapped due to a POLLHUP/POLLERR.
> +	 */
> +	if (!refcount_read(&map->refcnt))
> +		return -EINVAL;
> +
> +	*startp = overwrite ? head : old;
> +	*endp = overwrite ? old : head;
> +
> +	if (*startp == *endp)
> +		return -EAGAIN;
> +
> +	size = *endp - *startp;
> +	if (size > (unsigned long)(map->mask) + 1) {
> +		if (!overwrite) {
> +			WARN_ONCE(1, "failed to keep up with mmap data. (warn only once)\n");
> +
> +			map->prev = head;
> +			perf_mmap__consume(map, overwrite);
> +			return -EAGAIN;
> +		}
> +
> +		/*
> +		 * Backward ring buffer is full. We still have a chance to read
> +		 * most of data from it.
> +		 */
> +		if (overwrite_rb_find_range(data, map->mask, head, startp, endp))
> +			return -EINVAL;
> +	}
> +	return 0;
> +}
> +
>  int perf_mmap__push(struct perf_mmap *md, bool overwrite,
>  		    void *to, int push(void *to, void *buf, size_t size))
>  {
> diff --git a/tools/perf/util/mmap.h b/tools/perf/util/mmap.h
> index e43d7b5..0633308 100644
> --- a/tools/perf/util/mmap.h
> +++ b/tools/perf/util/mmap.h
> @@ -94,4 +94,6 @@ int perf_mmap__push(struct perf_mmap *md, bool backward,
>  
>  size_t perf_mmap__mmap_len(struct perf_mmap *map);
>  
> +int perf_mmap__read_init(struct perf_mmap *map, bool overwrite,
> +			 u64 *startp, u64 *endp);
>  #endif /*__PERF_MMAP_H */
> -- 
> 2.5.5
>