From: David Chinner <dgc@sgi.com>
To: xfs-dev <xfs-dev@sgi.com>
Cc: xfs-oss <xfs@oss.sgi.com>
Subject: [patch, debug, 1/2] use atomic counters for ktrace buffer indexes
Date: Tue, 19 Feb 2008 10:00:10 +1100 [thread overview]
Message-ID: <20080218230010.GT155407@sgi.com> (raw)
ktrace_enter() is consuming vast amounts of CPU time
due to the use of a single global lock for protecting
buffer index increments. Change it to use per-buffer
atomic counters - this reduces ktrace_enter() overhead
during a trace intensive test on a 4p machine from 58%
of all CPU time to 12% and halves test runtime.
Signed-off-by: Dave Chinner <dgc@sgi.com>
---
fs/xfs/support/ktrace.c | 21 ++++++++-------------
fs/xfs/support/ktrace.h | 2 +-
2 files changed, 9 insertions(+), 14 deletions(-)
Index: 2.6.x-xfs-new/fs/xfs/support/ktrace.c
===================================================================
--- 2.6.x-xfs-new.orig/fs/xfs/support/ktrace.c 2008-02-15 17:02:00.000000000 +1100
+++ 2.6.x-xfs-new/fs/xfs/support/ktrace.c 2008-02-18 15:33:13.849314747 +1100
@@ -92,7 +92,7 @@ ktrace_alloc(int nentries, unsigned int
ktp->kt_entries = ktep;
ktp->kt_nentries = nentries;
- ktp->kt_index = 0;
+ atomic_set(&ktp->kt_index, 0);
ktp->kt_rollover = 0;
return ktp;
}
@@ -151,8 +151,6 @@ ktrace_enter(
void *val14,
void *val15)
{
- static DEFINE_SPINLOCK(wrap_lock);
- unsigned long flags;
int index;
ktrace_entry_t *ktep;
@@ -161,12 +159,8 @@ ktrace_enter(
/*
* Grab an entry by pushing the index up to the next one.
*/
- spin_lock_irqsave(&wrap_lock, flags);
- index = ktp->kt_index;
- if (++ktp->kt_index == ktp->kt_nentries)
- ktp->kt_index = 0;
- spin_unlock_irqrestore(&wrap_lock, flags);
-
+ index = atomic_add_return(1, &ktp->kt_index);
+ index = (index - 1) % ktp->kt_nentries;
if (!ktp->kt_rollover && index == ktp->kt_nentries - 1)
ktp->kt_rollover = 1;
@@ -199,11 +193,12 @@ int
ktrace_nentries(
ktrace_t *ktp)
{
- if (ktp == NULL) {
+ int index;
+ if (ktp == NULL)
return 0;
- }
- return (ktp->kt_rollover ? ktp->kt_nentries : ktp->kt_index);
+ index = atomic_read(&ktp->kt_index) % ktp->kt_nentries;
+ return (ktp->kt_rollover ? ktp->kt_nentries : index);
}
/*
@@ -228,7 +223,7 @@ ktrace_first(ktrace_t *ktp, ktrace_sna
int nentries;
if (ktp->kt_rollover)
- index = ktp->kt_index;
+ index = atomic_read(&ktp->kt_index) % ktp->kt_nentries;
else
index = 0;
Index: 2.6.x-xfs-new/fs/xfs/support/ktrace.h
===================================================================
--- 2.6.x-xfs-new.orig/fs/xfs/support/ktrace.h 2007-10-02 16:01:47.000000000 +1000
+++ 2.6.x-xfs-new/fs/xfs/support/ktrace.h 2008-02-18 15:32:28.911073469 +1100
@@ -30,7 +30,7 @@ typedef struct ktrace_entry {
*/
typedef struct ktrace {
int kt_nentries; /* number of entries in trace buf */
- int kt_index; /* current index in entries */
+ atomic_t kt_index; /* current index in entries */
int kt_rollover;
ktrace_entry_t *kt_entries; /* buffer of entries */
} ktrace_t;
next reply other threads:[~2008-02-18 23:00 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-02-18 23:00 David Chinner [this message]
2008-02-22 4:05 ` [patch, debug, 1/2] use atomic counters for ktrace buffer indexes Christoph Hellwig
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=20080218230010.GT155407@sgi.com \
--to=dgc@sgi.com \
--cc=xfs-dev@sgi.com \
--cc=xfs@oss.sgi.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