git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Karsten Blees <karsten.blees@gmail.com>
To: Git List <git@vger.kernel.org>,
	msysGit <msysgit@googlegroups.com>,  Jeff King <peff@peff.net>
Subject: [PATCH v5 05/11] trace: add infrastructure to augment trace output with additional info
Date: Wed, 11 Jun 2014 09:59:13 +0200	[thread overview]
Message-ID: <53980C51.6020304@gmail.com> (raw)
In-Reply-To: <53980B83.9050409@gmail.com>

To be able to add a common prefix or suffix to all trace output (e.g.
a timestamp or file:line of the caller), factor out common setup and
cleanup tasks of the trace* functions.

When adding a common prefix, it makes sense that the output of each trace
call starts on a new line. Add '\n' in case the caller forgot.

Note that this explicitly limits trace output to line-by-line, it is no
longer possible to trace-print just part of a line. Until now, this was
just an implicit assumption (trace-printing part of a line worked, but
messed up the trace file if multiple threads or processes were involved).

Thread-safety / inter-process-safety is also the reason why we need to do
the prefixing and suffixing in memory rather than issuing multiple write()
calls. Write_or_whine_pipe() / xwrite() is atomic unless the size exceeds
MAX_IO_SIZE (8MB, see wrapper.c). In case of trace_strbuf, this costs an
additional string copy (which should be irrelevant for performance in light
of actual file IO).

While we're at it, rename trace_strbuf's 'buf' argument, which suggests
that the function is modifying the buffer. Trace_strbuf() currently is the
only trace API that can print arbitrary binary data (without barfing on
'%' or stopping at '\0'), so 'data' seems more appropriate.

Signed-off-by: Karsten Blees <blees@dcon.de>
---
 trace.c | 45 ++++++++++++++++++++++++++++++++++-----------
 trace.h |  2 +-
 2 files changed, 35 insertions(+), 12 deletions(-)

diff --git a/trace.c b/trace.c
index b7ca51b..c920429 100644
--- a/trace.c
+++ b/trace.c
@@ -77,17 +77,37 @@ static void do_trace_print(const char *key, const struct strbuf *buf)
 		close(fd);
 }
 
+static int prepare_trace_line(const char *key, struct strbuf *buf)
+{
+	if (!trace_want(key))
+		return 0;
+
+	set_try_to_free_routine(NULL);	/* is never reset */
+
+	/* add line prefix here */
+
+	return 1;
+}
+
+static void print_trace_line(const char *key, struct strbuf *buf)
+{
+	/* append newline if missing */
+	if (buf->len && buf->buf[buf->len - 1] != '\n')
+		strbuf_addch(buf, '\n');
+
+	do_trace_print(key, buf);
+	strbuf_release(buf);
+}
+
 static void trace_vprintf(const char *key, const char *format, va_list ap)
 {
 	struct strbuf buf = STRBUF_INIT;
 
-	if (!trace_want(key))
+	if (!prepare_trace_line(key, &buf))
 		return;
 
-	set_try_to_free_routine(NULL);	/* is never reset */
 	strbuf_vaddf(&buf, format, ap);
-	do_trace_print(key, &buf);
-	strbuf_release(&buf);
+	print_trace_line(key, &buf);
 }
 
 void trace_printf_key(const char *key, const char *format, ...)
@@ -106,9 +126,15 @@ void trace_printf(const char *format, ...)
 	va_end(ap);
 }
 
-void trace_strbuf(const char *key, const struct strbuf *buf)
+void trace_strbuf(const char *key, const struct strbuf *data)
 {
-	do_trace_print(key, buf);
+	struct strbuf buf = STRBUF_INIT;
+
+	if (!prepare_trace_line(key, &buf))
+		return;
+
+	strbuf_addbuf(&buf, data);
+	print_trace_line(key, &buf);
 }
 
 void trace_argv_printf(const char **argv, const char *format, ...)
@@ -116,18 +142,15 @@ void trace_argv_printf(const char **argv, const char *format, ...)
 	struct strbuf buf = STRBUF_INIT;
 	va_list ap;
 
-	if (!trace_want("GIT_TRACE"))
+	if (!prepare_trace_line("GIT_TRACE", &buf))
 		return;
 
-	set_try_to_free_routine(NULL);	/* is never reset */
 	va_start(ap, format);
 	strbuf_vaddf(&buf, format, ap);
 	va_end(ap);
 
 	sq_quote_argv(&buf, argv, 0);
-	strbuf_addch(&buf, '\n');
-	do_trace_print("GIT_TRACE", &buf);
-	strbuf_release(&buf);
+	print_trace_line("GIT_TRACE", &buf);
 }
 
 static const char *quote_crnl(const char *path)
diff --git a/trace.h b/trace.h
index 8fea50b..e03db2f 100644
--- a/trace.h
+++ b/trace.h
@@ -12,6 +12,6 @@ extern void trace_repo_setup(const char *prefix);
 extern int trace_want(const char *key);
 __attribute__((format (printf, 2, 3)))
 extern void trace_printf_key(const char *key, const char *format, ...);
-extern void trace_strbuf(const char *key, const struct strbuf *buf);
+extern void trace_strbuf(const char *key, const struct strbuf *data);
 
 #endif /* TRACE_H */
-- 
1.9.2.msysgit.0.501.gaeecf09

-- 
-- 
*** Please reply-to-all at all times ***
*** (do not pretend to know who is subscribed and who is not) ***
*** Please avoid top-posting. ***
The msysGit Wiki is here: https://github.com/msysgit/msysgit/wiki - Github accounts are free.

You received this message because you are subscribed to the Google
Groups "msysGit" group.
To post to this group, send email to msysgit@googlegroups.com
To unsubscribe from this group, send email to
msysgit+unsubscribe@googlegroups.com
For more options, and view previous threads, visit this group at
http://groups.google.com/group/msysgit?hl=en_US?hl=en

--- 
You received this message because you are subscribed to the Google Groups "msysGit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to msysgit+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

  parent reply	other threads:[~2014-06-11  7:59 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-06-11  7:55 [PATCH v5 00/11] add performance tracing facility Karsten Blees
2014-06-11  7:56 ` [PATCH v5 01/11] trace: move trace declarations from cache.h to new trace.h Karsten Blees
2014-06-11  7:57 ` [PATCH v5 02/11] trace: consistently name the format parameter Karsten Blees
2014-06-11  7:57 ` [PATCH v5 03/11] trace: remove redundant printf format attribute Karsten Blees
2014-06-11  7:58 ` [PATCH v5 04/11] trace: factor out printing to the trace file Karsten Blees
2014-06-11  7:59 ` Karsten Blees [this message]
2014-06-11  7:59 ` [PATCH v5 06/11] trace: add current timestamp to all trace output Karsten Blees
2014-06-11  8:00 ` [PATCH v5 07/11] trace: move code around, in preparation to file:line output Karsten Blees
2014-06-11  8:00 ` [PATCH v5 08/11] trace: add 'file:line' to all trace output Karsten Blees
2014-06-11  8:01 ` [PATCH v5 09/11] trace: add high resolution timer function to debug performance issues Karsten Blees
2014-06-11  8:12   ` Karsten Blees
2014-06-17 16:44     ` Junio C Hamano
2014-06-18 15:14       ` Karsten Blees
2014-06-11  8:01 ` [PATCH v5 10/11] trace: add trace_performance facility " Karsten Blees
2014-06-17 17:11   ` Junio C Hamano
2014-06-18 15:14     ` Karsten Blees
2014-06-18 17:28       ` Junio C Hamano
2014-06-11  8:02 ` [PATCH v5 11/11] git: add performance tracing for git's main() function to debug scripts Karsten Blees
2014-06-12 18:30 ` [PATCH v5 00/11] add performance tracing facility Junio C Hamano
2014-06-18 15:14   ` Karsten Blees
2014-06-25 14:28 ` Duy Nguyen
2014-06-25 14:49   ` Karsten Blees
2014-06-26  1:11     ` Duy Nguyen

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=53980C51.6020304@gmail.com \
    --to=karsten.blees@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=msysgit@googlegroups.com \
    --cc=peff@peff.net \
    /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;
as well as URLs for NNTP newsgroup(s).