Messages in this thread | | | Date | Tue, 07 Jan 2014 11:48:59 +0900 | From | Namhyung Kim <> | Subject | Re: [PATCH 1/4] tools lib traceevent: Add state member to struct trace_seq |
| |
Hi Steve and Jiri,
2014-01-06 PM 11:45, Steven Rostedt wrote: > On Mon, 6 Jan 2014 15:38:28 +0100 > Jiri Olsa <jolsa@redhat.com> wrote: > > >>> I just thought that it's not so important to print message so keeps the >>> error internally until it gets printed. But I can be wrong as usual... >> >> I think that if she manages to get one of those errors >> the perf would fail soon anyway.. so it feels better >> to print it out immediately. > > Yeah, using a trace_seq after it has been destroyed is a critical > failure, and a major bug. A print to the user console should not be a > problem here. And actually, crashing is not that bad either, as glibc > does the same with using free() of a freed pointer. > > But as this error is major, an unwanted print is minor.
OK, I'll add the WARN_ONCE in the TRACE_SEQ_CHECK then.
Thanks, Namhyung
| |