From: Masami Hiramatsu <mhiramat@kernel.org>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: LKML <linux-kernel@vger.kernel.org>, Ingo Molnar <mingo@kernel.org>
Subject: Re: [RFC PATCH v2 1/3] tracing: Show real address for trace event arguments
Date: Thu, 15 Oct 2020 22:51:37 +0900 [thread overview]
Message-ID: <20201015225137.cb9c4638b7f576ad96fe6105@kernel.org> (raw)
In-Reply-To: <20201015090955.5a48b0ba@gandalf.local.home>
On Thu, 15 Oct 2020 09:09:55 -0400
Steven Rostedt <rostedt@goodmis.org> wrote:
> On Thu, 15 Oct 2020 18:00:08 +0900
> Masami Hiramatsu <mhiramat@kernel.org> wrote:
>
> > +#define STATIC_FMT_BUF_SIZE 128
> > +static char static_fmt_buf[STATIC_FMT_BUF_SIZE];
> > +
> > +const char *trace_event_format(struct trace_iterator *iter, const char *fmt)
> > +{
> > + const char *p, *new_fmt;
> > + char *q;
> > +
> > + if (WARN_ON_ONCE(!fmt))
> > + return fmt;
> > +retry:
> > + p = fmt;
> > + new_fmt = q = iter->fmt;
> > + while (*p) {
> > + if (unlikely(q - new_fmt + 3 > iter->fmt_size)) {
> > + /* expand format buffer if needed */
> > + if (iter->fmt == static_fmt_buf)
> > + return fmt;
> > +
> > + iter->fmt_size = iter->fmt_size ? iter->fmt_size * 2
> > + : STATIC_FMT_BUF_SIZE;
>
> Doubling may be too much, as fmt strings are not going to grow
> exponentially.
OK, then I just make it "+= STATIC_FMT_BUF_SIZE".
Actually, as far as I can see, the format string should be shorter than
128 bytes. If we see it is longer than that, it will be 2 lines in event
definition. And 128 + 128 = 256, which will be longer than 3 lines in 80
columns wide terminal. I guess no one want to write it. :)
>
> > + kfree(iter->fmt);
> > + iter->fmt = kmalloc(iter->fmt_size, GFP_KERNEL);
>
> Need to test the return value of kmalloc. But I would use realloc instead,
> and continue:
Good catch!
>
> char *tmp_fmt;
> unsigned int tmp_size;
>
> tmp_size = iter->fmt_size + STATIC_FMT_BUF_SIZE;
>
> tmp_fmt = krealloc(iter->fmt, tmp_size, GFP_KERNEL);
> if (!tmp_fmt)
> return fmt;
>
> q += tmp_fmt - new_fmt;
> new_fmt = iter->fmt = tmp_fmt;
> iter->fmt_size = tmp_size;
>
> > + goto retry;
>
> Then you don't even need the retry, and just continue processing.
>
> Also, if it fails to allocate, you still have the iter->fmt and fmt_size
> intact and correct.
OK, let me update it.
Thank you,
>
> -- Steve
>
> > + }
> > + *q++ = *p++;
> > + /* Replace %p with %px */
> > + if (p[-1] == '%') {
> > + if (p[0] == '%') {
> > + *q++ = *p++;
> > + } else if (p[0] == 'p' && !isalnum(p[1])) {
> > + *q++ = *p++;
> > + *q++ = 'x';
> > + }
> > + }
> > + }
> > + *q = '\0';
> > +
> > + return new_fmt;
> > +}
> > +
--
Masami Hiramatsu <mhiramat@kernel.org>
next prev parent reply other threads:[~2020-10-15 13:51 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-10-15 8:59 [RFC PATCH v2 0/3] tracing: Show real address instead of hashed pointer Masami Hiramatsu
2020-10-15 9:00 ` [RFC PATCH v2 1/3] tracing: Show real address for trace event arguments Masami Hiramatsu
2020-10-15 13:09 ` Steven Rostedt
2020-10-15 13:51 ` Masami Hiramatsu [this message]
2020-10-15 9:00 ` [RFC PATCH v2 2/3] tracing: Update the stage 3 of trace event macro comment Masami Hiramatsu
2020-10-15 9:00 ` [RFC PATCH v2 3/3] tracing: Add ptr-hash option to show the hashed pointer value Masami Hiramatsu
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=20201015225137.cb9c4638b7f576ad96fe6105@kernel.org \
--to=mhiramat@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--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
all inboxes | Powered by JetHome®