Messages in this thread | | | Date | Tue, 10 Mar 2009 11:08:32 +0100 | From | Frederic Weisbecker <> | Subject | Re: [RFC][PATCH 1/2] tracing/ftrace: syscall tracing infrastructure |
| |
On Mon, Mar 09, 2009 at 08:52:00PM -0400, Steven Rostedt wrote: > > > On Mon, 9 Mar 2009, Frederic Weisbecker wrote: > > > On Mon, Mar 09, 2009 at 09:52:20AM -0400, Steven Rostedt wrote: > > > > > > On Sat, 7 Mar 2009, Frederic Weisbecker wrote: > > > > + > > > > + > > > > +static atomic_t refcount; > > > > + > > > > +enum print_line_t > > > > +print_syscall_enter(struct trace_iterator *iter, int flags) > > > > +{ > > > > + struct trace_seq *s = &iter->seq; > > > > + struct trace_entry *ent = iter->ent; > > > > + struct syscall_trace_enter *trace; > > > > + enum syscall_trace_nr syscall; > > > > + const struct syscall_trace_entry *entry; > > > > + char *str = NULL; > > > > + int i, ret; > > > > + > > > > + trace_assign_type(trace, ent); > > > > + > > > > + syscall = trace->nr; > > > > + > > > > + if (!syscall || syscall >= __SYSCALL_TRACE_END) > > > > > > You still need to print a "\n" otherwise the output of the next line > > > will start on the current line. > > > > Are you sure? > > This entry will be ignored. And the previous one did print a newline. > > So the next one should not have any problem. > > Does this take over the printline? Otherwise we could be printing a > header, (comm, timestamp, cpu, etc). > > -- Steve
Ah ok, I did not think of the headers.
| |