Messages in this thread |  | | | Date | Tue, 17 Mar 2009 11:41:05 -0400 (EDT) | | From | Steven Rostedt <> | | Subject | Re: [PATCH 4/7] tracing: new format for specialized trace points |
| |
On Tue, 17 Mar 2009, Christoph Hellwig wrote:
> On Tue, Mar 10, 2009 at 12:57:14AM -0400, Steven Rostedt wrote: > > Here's the example. The only updated macro in this patch is the > > sched_switch trace point. > > Note that we shouldn't keep two variants around long-term, that's > just going to cause confusion. > > > The old method looked like this: > > > > TRACE_EVENT_FORMAT(sched_switch, > > TP_PROTO(struct rq *rq, struct task_struct *prev, > > struct task_struct *next), > > TP_ARGS(rq, prev, next), > > TP_FMT("task %s:%d ==> %s:%d", > > prev->comm, prev->pid, next->comm, next->pid), > > TRACE_STRUCT( > > TRACE_FIELD(pid_t, prev_pid, prev->pid) > > TRACE_FIELD(int, prev_prio, prev->prio) > > TRACE_FIELD_SPECIAL(char next_comm[TASK_COMM_LEN], > > next_comm, > > TP_CMD(memcpy(TRACE_ENTRY->next_comm, > > next->comm, > > TASK_COMM_LEN))) > > TRACE_FIELD(pid_t, next_pid, next->pid) > > TRACE_FIELD(int, next_prio, next->prio) > > ), > > TP_RAW_FMT("prev %d:%d ==> next %s:%d:%d") > > ); > > > > The above method is hard to read and requires two format fields. > > > > The new method: > > > > /* > > * Tracepoint for task switches, performed by the scheduler: > > * > > * (NOTE: the 'rq' argument is not used by generic trace events, > > * but used by the latency tracer plugin. ) > > */ > > TRACE_EVENT(sched_switch, > > > > TP_PROTO(struct rq *rq, struct task_struct *prev, > > struct task_struct *next), > > > > TP_ARGS(rq, prev, next), > > > > TP_STRUCT__entry( > > __array( char, prev_comm, TASK_COMM_LEN ) > > __field( pid_t, prev_pid ) > > __field( int, prev_prio ) > > __array( char, next_comm, TASK_COMM_LEN ) > > __field( pid_t, next_pid ) > > __field( int, next_prio ) > > ), > > > > TP_printk("task %s:%d [%d] ==> %s:%d [%d]", > > __entry->prev_comm, __entry->prev_pid, __entry->prev_prio, > > __entry->next_comm, __entry->next_pid, __entry->next_prio), > > > > TP_fast_assign( > > memcpy(__entry->next_comm, next->comm, TASK_COMM_LEN); > > __entry->prev_pid = prev->pid; > > __entry->prev_prio = prev->prio; > > memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN); > > __entry->next_pid = next->pid; > > __entry->next_prio = next->prio; > > ) > > ); > > While the idea behing it seems like an improvement to me, the > implementation feel actually worse than the old one too me. I would > expect this to look more like: > > struct trace_sched_switch { > char prev_comm[TASK_COMM_LEN], > pid_t prev_pid, > int prev_prio, > char next_comm[TASK_COMM_LEN], > pid_t next_pid, > int next_prio, > }
We would love to do the above. The problem is that we also need a way to automatically export the fields offset/size to userspace. Thus we use the "__field()" and "__array()" macros to do this for us. Otherwise, we need to do that manually.
> > static void trace_sched_assign(struct trace_sched_switch *dst, struct rq *rq, > struct task_struct *prev, struct task_struct *next) > { > memcpy(dst->next_comm, next->comm, TASK_COMM_LEN); > dst->prev_pid = prev->pid; > dst->prev_prio = prev->prio; > memcpy(dst->prev_comm, prev->comm, TASK_COMM_LEN); > dst->next_pid = next->pid; > dst->next_prio = next->prio; > };
This we could take out of the macro and make a function.
-- Steve
> > > TRACE_EVENT(sched_switch, > trace_proto(struct rq *rq, struct task_struct *prev, > struct task_struct *next), > trace_args(rq, prev, next), > trace_struct(struct trace_sched_switch), > trace_assign(trace_sched_assign); > > trace_pretty_print("task %s:%d [%d] ==> %s:%d [%d]", > __entry->prev_comm, __entry->prev_pid, __entry->prev_prio, > __entry->next_comm, __entry->next_pid, __entry->next_prio), > ); > >
|  |