Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1621100

Re: [PATCH 2/7] perf inject: copy events when reordering events in pipe mode

From Jiri Olsa <jolsa@redhat.com>
Newsgroups linux.kernel
Subject Re: [PATCH 2/7] perf inject: copy events when reordering events in pipe mode
Date 2017-04-11 11:50 +0200
Message-ID <tv0AV-5dh-15@gated-at.bofh.it> (permalink)
References <tuNX3-5AE-3@gated-at.bofh.it> <tuNX4-5AE-43@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


On Mon, Apr 10, 2017 at 01:14:27PM -0700, David Carrillo-Cisneros wrote:
> __perf_session__process_pipe_events reuses the same memory buffer to
> process all events in the pipe.
> 
> When reordering is needed (e.g. -b option), events are not immediately
> flushed, but kept around until reordering is possible, causing
> memory corruption.
> 
> The problem is usually observed by a "Unknown sample error" output. It
> can easily be reproduced by:
> 
>   perf record -o - noploop | perf inject -b > output
> 
> Signed-off-by: David Carrillo-Cisneros <davidcc@google.com>
> ---
>  tools/perf/util/ordered-events.c | 3 ++-
>  tools/perf/util/session.c        | 1 +
>  2 files changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/perf/util/ordered-events.c b/tools/perf/util/ordered-events.c
> index fe84df1875aa..e70e935b1841 100644
> --- a/tools/perf/util/ordered-events.c
> +++ b/tools/perf/util/ordered-events.c
> @@ -79,7 +79,7 @@ static union perf_event *dup_event(struct ordered_events *oe,
>  
>  static void free_dup_event(struct ordered_events *oe, union perf_event *event)
>  {
> -	if (oe->copy_on_queue) {
> +	if (event && oe->copy_on_queue) {
>  		oe->cur_alloc_size -= event->header.size;
>  		free(event);
>  	}
> @@ -150,6 +150,7 @@ void ordered_events__delete(struct ordered_events *oe, struct ordered_event *eve
>  	list_move(&event->list, &oe->cache);
>  	oe->nr_events--;
>  	free_dup_event(oe, event->event);
> +	event->event = NULL;

based on the changelog I understand the need for the change below,
but I dont get why do we need this cleanu and check above

thanks,
jirka

>  }
>  
>  int ordered_events__queue(struct ordered_events *oe, union perf_event *event,
> diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
> index 24259bc2c598..a25302bc55a8 100644
> --- a/tools/perf/util/session.c
> +++ b/tools/perf/util/session.c
> @@ -1656,6 +1656,7 @@ static int __perf_session__process_pipe_events(struct perf_session *session)
>  	buf = malloc(cur_size);
>  	if (!buf)
>  		return -errno;
> +	ordered_events__set_copy_on_queue(oe, true);
>  more:
>  	event = buf;
>  	err = readn(fd, event, sizeof(struct perf_event_header));
> -- 
> 2.12.2.715.g7642488e1d-goog
> 

Back to linux.kernel | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

[PATCH 0/7] perf tool: pipe-mode fixes David Carrillo-Cisneros <davidcc@google.com> - 2017-04-10 22:20 +0200
  [PATCH 1/7] perf inject: don't proceed if perf_session__process_event fails David Carrillo-Cisneros <davidcc@google.com> - 2017-04-10 22:20 +0200
    [tip:perf/core] perf inject: Don't proceed if  perf_session__process_event() fails tip-bot for David Carrillo-Cisneros <tipbot@zytor.com> - 2017-04-12 07:50 +0200
  [PATCH 2/7] perf inject: copy events when reordering events in pipe mode David Carrillo-Cisneros <davidcc@google.com> - 2017-04-10 22:20 +0200
    Re: [PATCH 2/7] perf inject: copy events when reordering events in  pipe mode Jiri Olsa <jolsa@redhat.com> - 2017-04-11 11:50 +0200
      Re: [PATCH 2/7] perf inject: copy events when reordering events in  pipe mode David Carrillo-Cisneros <davidcc@google.com> - 2017-04-11 21:00 +0200
    [tip:perf/core] perf inject: Copy events when reordering events in  pipe mode tip-bot for David Carrillo-Cisneros <tipbot@zytor.com> - 2017-04-12 07:50 +0200
  [PATCH 4/7] perf annotate: process attr and build_id records David Carrillo-Cisneros <davidcc@google.com> - 2017-04-10 22:20 +0200
    [tip:perf/core] perf annotate: Process attr and build_id records tip-bot for David Carrillo-Cisneros <tipbot@zytor.com> - 2017-04-12 07:50 +0200
  [PATCH 6/7] perf tool: protect empty evlists David Carrillo-Cisneros <davidcc@google.com> - 2017-04-10 22:20 +0200
    Re: [PATCH 6/7] perf tool: protect empty evlists Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-04-11 20:20 +0200
      Re: [PATCH 6/7] perf tool: protect empty evlists David Carrillo-Cisneros <davidcc@google.com> - 2017-04-11 20:40 +0200
  [PATCH 7/7] perf tool: do not print missing features in pipe-mode David Carrillo-Cisneros <davidcc@google.com> - 2017-04-10 22:20 +0200
    [tip:perf/core] perf tools: Do not print missing features in  pipe-mode tip-bot for David Carrillo-Cisneros <tipbot@zytor.com> - 2017-04-12 07:50 +0200
  [PATCH 5/7] perf session: don't rely on evlist in pipe mode David Carrillo-Cisneros <davidcc@google.com> - 2017-04-10 22:20 +0200
    Re: [PATCH 5/7] perf session: don't rely on evlist in pipe mode Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-04-11 20:50 +0200
      Re: [PATCH 5/7] perf session: don't rely on evlist in pipe mode David Carrillo-Cisneros <davidcc@google.com> - 2017-04-11 21:00 +0200
        Re: [PATCH 5/7] perf session: don't rely on evlist in pipe mode Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-04-11 21:50 +0200
    Re: [PATCH 5/7] perf session: don't rely on evlist in pipe mode Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-04-11 20:50 +0200
    Re: [PATCH 5/7] perf session: don't rely on evlist in pipe mode Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-04-11 20:50 +0200
    [tip:perf/core] perf session: Don't rely on evlist in pipe mode tip-bot for David Carrillo-Cisneros <tipbot@zytor.com> - 2017-04-12 07:50 +0200
  Re: [PATCH 0/7] perf tool: pipe-mode fixes Jiri Olsa <jolsa@redhat.com> - 2017-04-11 11:50 +0200
    Re: [PATCH 0/7] perf tool: pipe-mode fixes Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-04-11 20:20 +0200

csiph-web