From: Tom Zanussi <tzanussi@gmail.com>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: linux-kernel <linux-kernel@vger.kernel.org>,
"Ingo Molnar" <mingo@elte.hu>,
"Frédéric Weisbecker" <fweisbec@gmail.com>
Subject: Re: [PATCH 3/4] tracing: add per-event filtering
Date: Tue, 24 Mar 2009 02:18:05 -0500 [thread overview]
Message-ID: <1237879085.8339.64.camel@charm-linux> (raw)
In-Reply-To: <alpine.DEB.2.00.0903231358320.3578@gandalf.stny.rr.com>
Hi,
On Mon, 2009-03-23 at 14:06 -0400, Steven Rostedt wrote:
> On Sun, 22 Mar 2009, Tom Zanussi wrote:
> >
> > +static ssize_t
> > +event_filter_read(struct file *filp, char __user *ubuf, size_t cnt,
> > + loff_t *ppos)
> > +{
> > + struct ftrace_event_call *call = filp->private_data;
> > + struct trace_seq *s;
> > + int r;
> > +
> > + if (*ppos)
> > + return 0;
> > +
> > + s = kmalloc(sizeof(*s), GFP_KERNEL);
> > + if (!s)
> > + return -ENOMEM;
> > +
> > + trace_seq_init(s);
> > +
> > + r = filter_print_preds(call->preds, s->buffer);
>
> You're not using any of the features of the trace_seq structure.
> Might as well just allocate your own buffer.
>
Yeah, it make more sense to use the trace_seq features than do it
myself. I just posted a patch to do that.
> > + r = simple_read_from_buffer(ubuf, cnt, ppos, s->buffer, r);
> > +
> > + kfree(s);
> > +
> > + return r;
> > +}
> > +
> > +static ssize_t
> > +event_filter_write(struct file *filp, const char __user *ubuf, size_t cnt,
> > + loff_t *ppos)
> > +{
> > + struct ftrace_event_call *call = filp->private_data;
> > + char buf[64], *pbuf = buf;
> > + struct filter_pred *pred;
> > + int err;
> > +
> > + if (cnt >= sizeof(buf))
> > + return -EINVAL;
> > +
> > + if (copy_from_user(&buf, ubuf, cnt))
> > + return -EFAULT;
> > +
> > + pred = kzalloc(sizeof(*pred), GFP_KERNEL);
> > + if (!pred)
> > + return -ENOMEM;
> > +
> > + err = filter_parse(&pbuf, pred);
> > + if (err < 0) {
> > + filter_free_pred(pred);
> > + return err;
> > + }
> > +
> > + if (pred->clear) {
> > + filter_free_preds(call);
>
> The above is very confusing. Why are we passing in "call"?
The preds are attached to the call, so it makes sense to me to pass in
the call. I could just pass in the preds directly, if that makes it
less confusing...
>
> > + return cnt;
> > + }
> > +
> > + if (filter_add_pred(call, pred)) {
> > + filter_free_pred(pred);
> > + return -EINVAL;
> > + }
> > +
> > + *ppos += cnt;
> > +
> > + return cnt;
> > +}
> > +
> > static const struct seq_operations show_event_seq_ops = {
> > .start = t_start,
> > .next = t_next,
> > @@ -504,6 +569,12 @@ static const struct file_operations ftrace_event_id_fops = {
> > .read = event_id_read,
> > };
> >
> > +static const struct file_operations ftrace_event_filter_fops = {
> > + .open = tracing_open_generic,
> > + .read = event_filter_read,
> > + .write = event_filter_write,
> > +};
> > +
> > static struct dentry *event_trace_events_dir(void)
> > {
> > static struct dentry *d_tracer;
> > @@ -619,6 +690,12 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events)
> > }
> > }
> >
> > + entry = debugfs_create_file("filter", 0444, call->dir, call,
> > + &ftrace_event_filter_fops);
> > + if (!entry)
> > + pr_warning("Could not create debugfs "
> > + "'%s/filter' entry\n", call->name);
> > +
> > /* A trace may not want to export its format */
> > if (!call->show_format)
> > return 0;
>
> [...]
>
> > +
> > diff --git a/kernel/trace/trace_events_stage_3.h b/kernel/trace/trace_events_stage_3.h
> > index 468938f..ebf215e 100644
> > --- a/kernel/trace/trace_events_stage_3.h
> > +++ b/kernel/trace/trace_events_stage_3.h
> > @@ -204,6 +204,7 @@ static struct ftrace_event_call event_##call; \
> > \
> > static void ftrace_raw_event_##call(proto) \
> > { \
> > + struct ftrace_event_call *call = &event_##call; \
> > struct ring_buffer_event *event; \
> > struct ftrace_raw_##call *entry; \
> > unsigned long irq_flags; \
> > @@ -222,6 +223,9 @@ static void ftrace_raw_event_##call(proto) \
> > assign; \
> > \
> > trace_current_buffer_unlock_commit(event, irq_flags, pc); \
> > + \
> > + if (call->preds && !filter_match_preds(call, entry)) \
> > + ring_buffer_event_discard(event); \
> > } \
>
> Again, this would need to be within the commit.
>
Yeah, Frédéric already posted a patch to do that.
Thanks for your comments!
Tom
> -- Steve
>
> > \
> > static int ftrace_raw_reg_event_##call(void) \
> > --
> > 1.5.6.3
> >
> >
> >
> >
> >
next prev parent reply other threads:[~2009-03-24 7:18 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-22 8:31 [PATCH 3/4] tracing: add per-event filtering Tom Zanussi
2009-03-22 18:48 ` Frederic Weisbecker
2009-03-22 19:40 ` [tip:tracing/filters] " Tom Zanussi
2009-03-23 18:06 ` [PATCH 3/4] " Steven Rostedt
2009-03-24 7:18 ` Tom Zanussi [this message]
2009-03-24 16:01 ` Steven Rostedt
2009-03-25 5:23 ` Tom Zanussi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1237879085.8339.64.camel@charm-linux \
--to=tzanussi@gmail.com \
--cc=fweisbec@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=rostedt@goodmis.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox