mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: linux-tip-commits@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, paulus@samba.org, hpa@zytor.com,
	mingo@redhat.com, a.p.zijlstra@chello.nl, tglx@linutronix.de,
	mingo@elte.hu
Subject: [tip:perfcounters/core] perf_counter: provide generic callchain bits
Date: Wed, 1 Apr 2009 10:14:46 GMT	[thread overview]
Message-ID: <tip-1254256f65ce3973462dc153e02b4f3a7e314e20@git.kernel.org> (raw)
In-Reply-To: <20090330171024.254266860@chello.nl>

Commit-ID:  1254256f65ce3973462dc153e02b4f3a7e314e20
Gitweb:     http://git.kernel.org/tip/1254256f65ce3973462dc153e02b4f3a7e314e20
Author:     Peter Zijlstra <a.p.zijlstra@chello.nl>
AuthorDate: Mon, 30 Mar 2009 19:07:14 +0200
Committer:  Ingo Molnar <mingo@elte.hu>
CommitDate: Wed, 1 Apr 2009 11:33:38 +0200

perf_counter: provide generic callchain bits

Provide the generic callchain support bits. If hw_event->callchain is
set the arch specific perf_callchain() function is called upon to
provide a perf_callchain_entry structure filled with the current
callchain.

If it does so, it is added to the overflow output event.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Paul Mackerras <paulus@samba.org>
LKML-Reference: <20090330171024.254266860@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>


---
 include/linux/perf_counter.h |   13 ++++++++++++-
 kernel/perf_counter.c        |   27 +++++++++++++++++++++++++++
 2 files changed, 39 insertions(+), 1 deletions(-)

diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h
index edf5bfb..43083af 100644
--- a/include/linux/perf_counter.h
+++ b/include/linux/perf_counter.h
@@ -140,8 +140,9 @@ struct perf_counter_hw_event {
 				include_tid    :  1, /* include the tid       */
 				mmap           :  1, /* include mmap data     */
 				munmap         :  1, /* include munmap data   */
+				callchain      :  1, /* add callchain data    */
 
-				__reserved_1   : 52;
+				__reserved_1   : 51;
 
 	__u32			extra_config_len;
 	__u32			__reserved_4;
@@ -219,6 +220,7 @@ enum perf_event_type {
 	PERF_EVENT_OVERFLOW	= 1UL << 31,
 	__PERF_EVENT_IP		= 1UL << 30,
 	__PERF_EVENT_TID	= 1UL << 29,
+	__PERF_EVENT_CALLCHAIN  = 1UL << 28,
 };
 
 #ifdef __KERNEL__
@@ -504,6 +506,15 @@ extern void perf_counter_mmap(unsigned long addr, unsigned long len,
 extern void perf_counter_munmap(unsigned long addr, unsigned long len,
 				unsigned long pgoff, struct file *file);
 
+#define MAX_STACK_DEPTH		255
+
+struct perf_callchain_entry {
+	u64	nr;
+	u64	ip[MAX_STACK_DEPTH];
+};
+
+extern struct perf_callchain_entry *perf_callchain(struct pt_regs *regs);
+
 #else
 static inline void
 perf_counter_task_sched_in(struct task_struct *task, int cpu)		{ }
diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c
index d93e9dd..860cdc2 100644
--- a/kernel/perf_counter.c
+++ b/kernel/perf_counter.c
@@ -1654,6 +1654,17 @@ void perf_counter_do_pending(void)
 }
 
 /*
+ * Callchain support -- arch specific
+ */
+
+struct perf_callchain_entry *
+__attribute__((weak))
+perf_callchain(struct pt_regs *regs)
+{
+	return NULL;
+}
+
+/*
  * Output
  */
 
@@ -1764,6 +1775,8 @@ static void perf_output_simple(struct perf_counter *counter,
 	struct {
 		u32 pid, tid;
 	} tid_entry;
+	struct perf_callchain_entry *callchain = NULL;
+	int callchain_size = 0;
 
 	header.type = PERF_EVENT_OVERFLOW;
 	header.size = sizeof(header);
@@ -1781,6 +1794,17 @@ static void perf_output_simple(struct perf_counter *counter,
 		header.size += sizeof(tid_entry);
 	}
 
+	if (counter->hw_event.callchain) {
+		callchain = perf_callchain(regs);
+
+		if (callchain) {
+			callchain_size = (1 + callchain->nr) * sizeof(u64);
+
+			header.type |= __PERF_EVENT_CALLCHAIN;
+			header.size += callchain_size;
+		}
+	}
+
 	ret = perf_output_begin(&handle, counter, header.size, nmi);
 	if (ret)
 		return;
@@ -1791,6 +1815,9 @@ static void perf_output_simple(struct perf_counter *counter,
 	if (counter->hw_event.include_tid)
 		perf_output_put(&handle, tid_entry);
 
+	if (callchain)
+		perf_output_copy(&handle, callchain, callchain_size);
+
 	perf_output_end(&handle);
 }
 

  parent reply	other threads:[~2009-04-01 10:19 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-30 17:07 [PATCH 00/15] pending perf_counter bits Peter Zijlstra
2009-03-30 17:07 ` [PATCH 01/15] perf_counter: unify and fix delayed counter wakeup Peter Zijlstra
2009-03-31  5:45   ` Paul Mackerras
2009-03-31  6:37     ` Peter Zijlstra
2009-04-01  9:04       ` Ingo Molnar
2009-04-01 10:12   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 02/15] perf_counter: fix update_userpage() Peter Zijlstra
2009-04-01 10:12   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 03/15] perf_counter: kerneltop: simplify data_head read Peter Zijlstra
2009-04-01 10:13   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 04/15] perf_counter: executable mmap() information Peter Zijlstra
2009-04-01 10:13   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 05/15] perf_counter: kerneltop: parse the mmap data stream Peter Zijlstra
2009-04-01 10:13   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 06/15] perf_counter: powerpc: only reserve PMU hardware when we need it Peter Zijlstra
2009-04-01 10:13   ` [tip:perfcounters/core] " Paul Mackerras
2009-03-30 17:07 ` [PATCH 07/15] perf_counter: make it possible for hw_perf_counter_init to return error codes Peter Zijlstra
2009-04-01 10:13   ` [tip:perfcounters/core] " Paul Mackerras
2009-03-30 17:07 ` [PATCH 08/15] perf_counter: x86: proper error propagation for the x86 hw_perf_counter_init() Peter Zijlstra
2009-04-01 10:13   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 09/15] perf_counter tools: optionally scale counter values in perfstat mode Peter Zijlstra
2009-04-01  9:39   ` Ingo Molnar
2009-04-01 10:50     ` Paul Mackerras
2009-04-01 10:14   ` [tip:perfcounters/core] " Paul Mackerras
2009-03-30 17:07 ` [PATCH 10/15] perf_counter: small cleanup of the output routines Peter Zijlstra
2009-04-01 10:14   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 11/15] perf_counter: re-arrange the perf_event_type Peter Zijlstra
2009-04-01 10:14   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 12/15] pref_counter: kerneltop: update event_types Peter Zijlstra
2009-04-01 10:14   ` [tip:perfcounters/core] perf_counter tools: " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 13/15] perf_counter: provide generic callchain bits Peter Zijlstra
2009-03-31  6:12   ` Paul Mackerras
2009-03-31  6:39     ` Peter Zijlstra
2009-03-31  7:24       ` Corey Ashford
2009-03-31  8:43         ` Peter Zijlstra
2009-03-31  9:09           ` Paul Mackerras
2009-03-31  9:12         ` Paul Mackerras
2009-03-31 14:00           ` Peter Zijlstra
2009-03-31 17:11             ` Corey Ashford
2009-04-01  3:48             ` Paul Mackerras
2009-04-01  7:59               ` Peter Zijlstra
2009-04-01  8:45                 ` Paul Mackerras
2009-04-01 10:00                   ` Ingo Molnar
2009-04-01 11:53                     ` Paul Mackerras
2009-04-01 23:25                 ` Corey Ashford
2009-04-02  6:43                   ` Peter Zijlstra
2009-04-02  7:41                     ` Peter Zijlstra
2009-04-02  9:10                       ` Paul Mackerras
2009-04-02  9:14                         ` Peter Zijlstra
2009-04-01 10:14   ` Peter Zijlstra [this message]
2009-03-30 17:07 ` [PATCH 14/15] perf_counter: x86: callchain support Peter Zijlstra
2009-04-01 10:14   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-30 17:07 ` [PATCH 15/15] perf_counter: pmc arbitration Peter Zijlstra
2009-04-01 10:15   ` [tip:perfcounters/core] " Peter Zijlstra
2009-03-31  5:43 ` [PATCH 00/15] pending perf_counter bits Paul Mackerras

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=tip-1254256f65ce3973462dc153e02b4f3a7e314e20@git.kernel.org \
    --to=a.p.zijlstra@chello.nl \
    --cc=hpa@zytor.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-tip-commits@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=mingo@redhat.com \
    --cc=paulus@samba.org \
    --cc=tglx@linutronix.de \
    /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

Powered by JetHome