From: Jianfeng Wang <jianfeng.w.wang@oracle.com>
To: linux-trace-devel@vger.kernel.org
Cc: rostedt@goodmis.org, jianfeng.w.wang@oracle.com
Subject: [PATCH 1/2] trace-cmd ftrace: print function retval in function_graph
Date: Mon, 15 Apr 2024 08:49:20 -0700 [thread overview]
Message-ID: <20240415154921.4998-2-jianfeng.w.wang@oracle.com> (raw)
In-Reply-To: <20240415154921.4998-1-jianfeng.w.wang@oracle.com>
The upstream Linux kernel has introduced a new feature, i.e.
funcgraph-retval for the function_graph tracer. (Commit ID:
a1be9ccc57f07d54278be34eed6bd679bc941c97). With this feature, the
function_graph tracer can record each function's return value along
with its execution time at the function_graph exit handler. This is
useful for debugging kernel issues, such as syscall errors.
This commit enhances 'trace-cmd report' by supporting the
function_graph retval feature if the underlying kernel supports it.
Once the feature is supported, 'trace-cmd report' will print each
function's return value in this format: '(ret=retval)" at the end
of every funcgraph leaf entry and exit event.
Signed-off-by: Jianfeng Wang <jianfeng.w.wang@oracle.com>
---
lib/trace-cmd/trace-ftrace.c | 26 ++++++++++++++++++++++++++
1 file changed, 26 insertions(+)
diff --git a/lib/trace-cmd/trace-ftrace.c b/lib/trace-cmd/trace-ftrace.c
index 89e46c3d..cb05d88c 100644
--- a/lib/trace-cmd/trace-ftrace.c
+++ b/lib/trace-cmd/trace-ftrace.c
@@ -185,6 +185,8 @@ print_graph_entry_leaf(struct trace_seq *s,
unsigned long long rettime, calltime;
unsigned long long duration, depth;
unsigned long long val;
+ unsigned long long retval;
+ bool fgraph_retval_supported = true;
const char *func;
int ret;
int i;
@@ -195,6 +197,13 @@ print_graph_entry_leaf(struct trace_seq *s,
if (tep_get_field_val(s, finfo->fgraph_ret_event, "calltime", ret_rec, &calltime, 1))
return trace_seq_putc(s, '!');
+ if (!tep_find_field(finfo->fgraph_ret_event, "retval"))
+ fgraph_retval_supported = false;
+ else {
+ if (tep_get_field_val(s, finfo->fgraph_ret_event, "retval", ret_rec, &retval, 1))
+ return trace_seq_putc(s, '!');
+ }
+
duration = rettime - calltime;
/* Overhead */
@@ -222,6 +231,10 @@ print_graph_entry_leaf(struct trace_seq *s,
if (ret && fgraph_depth->set)
ret = trace_seq_printf(s, " (%lld)", depth);
+ /* Return Value */
+ if (ret && fgraph_retval_supported)
+ ret = trace_seq_printf(s, " (ret=%lld)", retval);
+
return ret;
}
@@ -316,6 +329,8 @@ fgraph_ret_handler(struct trace_seq *s, struct tep_record *record,
unsigned long long duration, depth;
unsigned long long val;
const char *func;
+ unsigned long long retval;
+ bool fgraph_retval_supported = true;
int i;
ret_event_check(finfo, event->tep);
@@ -326,6 +341,13 @@ fgraph_ret_handler(struct trace_seq *s, struct tep_record *record,
if (tep_get_field_val(s, event, "calltime", record, &calltime, 1))
return trace_seq_putc(s, '!');
+ if (!tep_find_field(event, "retval"))
+ fgraph_retval_supported = false;
+ else {
+ if (tep_get_field_val(s, event, "retval", record, &retval, 1))
+ return trace_seq_putc(s, '!');
+ }
+
duration = rettime - calltime;
/* Overhead */
@@ -355,6 +377,10 @@ fgraph_ret_handler(struct trace_seq *s, struct tep_record *record,
if (fgraph_depth->set)
trace_seq_printf(s, " (%lld)", depth);
+ /* Return Value */
+ if (fgraph_retval_supported)
+ trace_seq_printf(s, " (ret=%lld)", retval);
+
return 0;
}
--
2.42.1
next prev parent reply other threads:[~2024-04-15 15:49 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-15 15:49 [PATCH 0/2] trace-cmd ftrace: support function retval feature in function_graph Jianfeng Wang
2024-04-15 15:49 ` Jianfeng Wang [this message]
2024-04-15 15:49 ` [PATCH 2/2] trace-cmd: Add ftrace options with fgraph retval option Jianfeng Wang
2024-04-30 22:45 ` Jianfeng Wang
2024-04-30 23:06 ` Steven Rostedt
2024-05-17 0:53 ` Steven Rostedt
2024-05-17 18:50 ` Jianfeng Wang
2024-05-17 19:00 ` Steven Rostedt
2024-05-17 19:23 ` [External] : " Jianfeng Wang
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=20240415154921.4998-2-jianfeng.w.wang@oracle.com \
--to=jianfeng.w.wang@oracle.com \
--cc=linux-trace-devel@vger.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).