From: Byungchul Park <byungchul.park@lge.com>
To: peterz@infradead.org, mingo@kernel.org
Cc: linux-kernel@vger.kernel.org, npiggin@suse.de, walken@google.com,
ak@suse.de, tglx@inhelltoy.tec.linutronix.de
Subject: [RFC 12/12] x86/dumpstack: Optimize save_stack_trace
Date: Mon, 20 Jun 2016 13:55:27 +0900 [thread overview]
Message-ID: <1466398527-1122-13-git-send-email-byungchul.park@lge.com> (raw)
In-Reply-To: <1466398527-1122-1-git-send-email-byungchul.park@lge.com>
Currently, x86 implementation of save_stack_trace() is walking all stack
region word by word regardless of what the trace->max_entries is.
However, it's unnecessary to walk after already fulfilling caller's
requirement, say, if trace->nr_entries >= trace->max_entries is true.
For example, CONFIG_LOCKDEP_CROSSRELEASE implementation calls
save_stack_trace() with max_entries = 5 frequently. I measured its
overhead and printed its difference of sched_clock() with my QEMU x86
machine.
The latency was improved over 70% when trace->max_entries = 5.
Before this patch:
[ 2.326940] save_stack_trace() takes 83931 ns
[ 2.326389] save_stack_trace() takes 62576 ns
[ 2.327575] save_stack_trace() takes 58826 ns
[ 2.327000] save_stack_trace() takes 88980 ns
[ 2.327424] save_stack_trace() takes 59831 ns
[ 2.327575] save_stack_trace() takes 58482 ns
[ 2.327597] save_stack_trace() takes 87114 ns
[ 2.327931] save_stack_trace() takes 121140 ns
[ 2.327434] save_stack_trace() takes 64321 ns
[ 2.328632] save_stack_trace() takes 84997 ns
[ 2.328000] save_stack_trace() takes 115037 ns
[ 2.328460] save_stack_trace() takes 72292 ns
[ 2.328632] save_stack_trace() takes 61236 ns
[ 2.328567] save_stack_trace() takes 76666 ns
[ 2.328867] save_stack_trace() takes 79525 ns
[ 2.328460] save_stack_trace() takes 64902 ns
[ 2.329585] save_stack_trace() takes 58760 ns
[ 2.329000] save_stack_trace() takes 91349 ns
[ 2.329414] save_stack_trace() takes 60069 ns
[ 2.329585] save_stack_trace() takes 61012 ns
[ 2.329573] save_stack_trace() takes 76820 ns
[ 2.329863] save_stack_trace() takes 62131 ns
[ 2.330000] save_stack_trace() takes 99476 ns
[ 2.329846] save_stack_trace() takes 62419 ns
[ 2.330000] save_stack_trace() takes 88918 ns
[ 2.330253] save_stack_trace() takes 73669 ns
[ 2.330520] save_stack_trace() takes 67876 ns
[ 2.330671] save_stack_trace() takes 75963 ns
[ 2.330983] save_stack_trace() takes 95079 ns
[ 2.330451] save_stack_trace() takes 62352 ns
After this patch:
[ 2.780735] save_stack_trace() takes 19902 ns
[ 2.780718] save_stack_trace() takes 20240 ns
[ 2.781692] save_stack_trace() takes 45215 ns
[ 2.781477] save_stack_trace() takes 20191 ns
[ 2.781694] save_stack_trace() takes 20044 ns
[ 2.782589] save_stack_trace() takes 20292 ns
[ 2.782706] save_stack_trace() takes 20024 ns
[ 2.782706] save_stack_trace() takes 19881 ns
[ 2.782881] save_stack_trace() takes 24577 ns
[ 2.782706] save_stack_trace() takes 19901 ns
[ 2.783621] save_stack_trace() takes 24381 ns
[ 2.783621] save_stack_trace() takes 20205 ns
[ 2.783760] save_stack_trace() takes 19956 ns
[ 2.783718] save_stack_trace() takes 20280 ns
[ 2.784179] save_stack_trace() takes 20099 ns
[ 2.784835] save_stack_trace() takes 20055 ns
[ 2.785922] save_stack_trace() takes 20157 ns
[ 2.785922] save_stack_trace() takes 20140 ns
[ 2.786178] save_stack_trace() takes 20040 ns
[ 2.786877] save_stack_trace() takes 20102 ns
[ 2.795000] save_stack_trace() takes 21147 ns
[ 2.795397] save_stack_trace() takes 20230 ns
[ 2.795397] save_stack_trace() takes 31274 ns
[ 2.795739] save_stack_trace() takes 19706 ns
[ 2.796484] save_stack_trace() takes 20266 ns
[ 2.796484] save_stack_trace() takes 20902 ns
[ 2.797000] save_stack_trace() takes 38110 ns
[ 2.797510] save_stack_trace() takes 20224 ns
[ 2.798181] save_stack_trace() takes 20172 ns
[ 2.798837] save_stack_trace() takes 20824 ns
Signed-off-by: Byungchul Park <byungchul.park@lge.com>
---
arch/x86/include/asm/stacktrace.h | 1 +
arch/x86/kernel/dumpstack.c | 2 ++
arch/x86/kernel/dumpstack_32.c | 2 ++
arch/x86/kernel/stacktrace.c | 7 +++++++
4 files changed, 12 insertions(+)
diff --git a/arch/x86/include/asm/stacktrace.h b/arch/x86/include/asm/stacktrace.h
index 70bbe39..fc572e7 100644
--- a/arch/x86/include/asm/stacktrace.h
+++ b/arch/x86/include/asm/stacktrace.h
@@ -41,6 +41,7 @@ struct stacktrace_ops {
/* On negative return stop dumping */
int (*stack)(void *data, char *name);
walk_stack_t walk_stack;
+ int (*end_walk)(void *data);
};
void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c
index 9c30acf..355fe8f 100644
--- a/arch/x86/kernel/dumpstack.c
+++ b/arch/x86/kernel/dumpstack.c
@@ -115,6 +115,8 @@ print_context_stack(struct thread_info *tinfo,
print_ftrace_graph_addr(addr, data, ops, tinfo, graph);
}
stack++;
+ if (ops->end_walk && ops->end_walk(data))
+ break;
}
return bp;
}
diff --git a/arch/x86/kernel/dumpstack_32.c b/arch/x86/kernel/dumpstack_32.c
index 464ffd6..cc51419 100644
--- a/arch/x86/kernel/dumpstack_32.c
+++ b/arch/x86/kernel/dumpstack_32.c
@@ -71,6 +71,8 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs,
context = task_thread_info(task);
bp = ops->walk_stack(context, stack, bp, ops, data,
end_stack, &graph);
+ if (ops->end_walk && ops->end_walk(data))
+ break;
/* Stop if not on irq stack */
if (!end_stack)
diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c
index fdd0c64..9545719 100644
--- a/arch/x86/kernel/stacktrace.c
+++ b/arch/x86/kernel/stacktrace.c
@@ -43,10 +43,17 @@ save_stack_address_nosched(void *data, unsigned long addr, int reliable)
return __save_stack_address(data, addr, reliable, true);
}
+static int save_stack_end(void *data)
+{
+ struct stack_trace *trace = data;
+ return trace->nr_entries >= trace->max_entries;
+}
+
static const struct stacktrace_ops save_stack_ops = {
.stack = save_stack_stack,
.address = save_stack_address,
.walk_stack = print_context_stack,
+ .end_walk = save_stack_end,
};
static const struct stacktrace_ops save_stack_ops_nosched = {
--
1.9.1
next prev parent reply other threads:[~2016-06-20 5:13 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-20 4:55 [RFC 00/12] lockdep: Implement crossrelease feature Byungchul Park
2016-06-20 4:55 ` [RFC 01/12] lockdep: Refactor lookup_chain_cache() Byungchul Park
2016-06-20 4:55 ` [RFC 02/12] lockdep: Add a function building a chain between two hlocks Byungchul Park
2016-06-20 4:55 ` [RFC 03/12] lockdep: Make check_prev_add can use a stack_trace of other context Byungchul Park
2016-06-20 4:55 ` [RFC 04/12] lockdep: Make save_trace can copy from other stack_trace Byungchul Park
2016-06-20 4:55 ` [RFC 05/12] lockdep: Implement crossrelease feature Byungchul Park
2016-06-30 13:03 ` Peter Zijlstra
2016-06-30 23:28 ` Byungchul Park
2016-06-20 4:55 ` [RFC 06/12] lockdep: Apply crossrelease to completion Byungchul Park
2016-06-20 4:55 ` [RFC 07/12] pagemap.h: Remove trailing white space Byungchul Park
2016-06-20 4:55 ` [RFC 08/12] lockdep: Apply crossrelease to PG_locked lock Byungchul Park
2016-06-30 13:04 ` Peter Zijlstra
2016-06-30 23:21 ` Byungchul Park
2016-07-01 8:10 ` Peter Zijlstra
2016-07-01 11:18 ` Kirill A. Shutemov
2016-07-04 4:30 ` Byungchul Park
2016-06-20 4:55 ` [RFC 09/12] cifs/file.c: Remove trailing white space Byungchul Park
2016-06-20 4:55 ` [RFC 10/12] mm/swap_state.c: " Byungchul Park
2016-06-20 4:55 ` [RFC 11/12] lockdep: Call lock_acquire(release) when accessing PG_locked manually Byungchul Park
2016-06-20 4:55 ` Byungchul Park [this message]
2016-06-20 7:29 ` [RFC 12/12] x86/dumpstack: Optimize save_stack_trace xinhui
2016-06-20 7:50 ` byungchul.park
2016-06-29 12:43 ` Byungchul Park
2016-06-30 10:38 ` xinhui
2016-06-30 23:06 ` Byungchul Park
2016-06-23 23:37 ` [RFC 00/12] lockdep: Implement crossrelease feature Byungchul Park
2016-06-24 7:08 ` Peter Zijlstra
2016-06-24 11:13 ` Byungchul Park
2016-06-24 11:26 ` Nikolay Borisov
2016-06-27 1:34 ` Byungchul Park
2016-07-01 4:15 ` [PATCH] lockdep: Add a document describing " Byungchul Park
2016-07-01 10:45 ` Peter Zijlstra
2016-07-04 6:42 ` Byungchul Park
2016-07-06 0:49 ` Boqun Feng
2016-07-06 2:17 ` Byungchul Park
2016-07-06 5:33 ` Byungchul Park
2016-07-06 7:56 ` Peter Zijlstra
2016-07-06 8:12 ` Byungchul Park
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=1466398527-1122-13-git-send-email-byungchul.park@lge.com \
--to=byungchul.park@lge.com \
--cc=ak@suse.de \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=npiggin@suse.de \
--cc=peterz@infradead.org \
--cc=tglx@inhelltoy.tec.linutronix.de \
--cc=walken@google.com \
/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).