Peter Zijlstra wrote:
On Wed, 2007-06-06 at 23:34 +0200, Martin Peschke wrote:
+#ifdef CONFIG_LOCK_STAT
+enum lock_stat_enum {
+ LOCK_STAT_CONT,
+ LOCK_STAT_WAIT_READ,
+ LOCK_STAT_WAIT_WRITE,
+ LOCK_STAT_HOLD_READ,
+ LOCK_STAT_HOLD_WRITE,
+ _LOCK_STAT_NUMBER
Bill Huey (hui) wrote:
> However, I don't understand why all of this
> is so heavy weight when the current measurements that Peter makes is
> completely sufficient for any reasonable purpose I can think of at the
> moment.
It wasn't meant to be or to stay heavy weight. To make the best of my
rece
Peter Zijlstra wrote:
On Fri, 2007-06-08 at 19:37 +0200, Martin Peschke wrote:
Peter Zijlstra wrote:
On Fri, 2007-06-08 at 19:18 +0200, Martin Peschke wrote:
Peter Zijlstra wrote:
On Fri, 2007-06-08 at 19:00 +0200, Martin Peschke wrote:
Peter Zijlstra wrote:
I'm confused as to wher
Peter Zijlstra wrote:
On Fri, 2007-06-08 at 19:18 +0200, Martin Peschke wrote:
Peter Zijlstra wrote:
On Fri, 2007-06-08 at 19:00 +0200, Martin Peschke wrote:
Peter Zijlstra wrote:
I'm confused as to where the class->stat objects are initialised? Is
that done in lock_stat_init()? If
Peter Zijlstra wrote:
On Wed, 2007-06-06 at 23:34 +0200, Martin Peschke wrote:
+struct statistic_info lock_stat_info[_LOCK_STAT_NUMBER] = {
+ [LOCK_STAT_CONT] = {
+ .name = "contentions",
+ .x_unit = "instruction_pointer",
+
Peter Zijlstra wrote:
I'm confused as to where the class->stat objects are initialised? Is
that done in lock_stat_init()? If so, then you have a bug.
static struct lock_class lock_classes[MAX_LOCKDEP_KEYS];
I assume this gets us class structures containing all zeros, doesn't it?
Then class->st
Ingo Molnar wrote:
> Firstly, submit cleanup patches that _do not change the output_. If you
> have any output changes, do it as a separate patch, ontop of the cleanup
> patch. Mixing material changes and cleanups into a single patch is a
> basic patch submission mistake that will only earn you NA
Ingo Molnar wrote:
if the infrastructure your are advocating does not allow us to keep the
existing output then it's simply not flexible enough.
Let's be precise. If "keep the existing output" means any format change is
unacceptible to you, then I broke things. If it means that my method provid
Bill Huey (hui) wrote:
> There are roughly about 400 locks in a normal kernel for a desktop. The
> list is rather cumbersome anyways so, IMO, it really should be handled
> by parsing tools, etc... There could be more properties attached to each
> lock especially if you intend to get this to work o
Peter Zijlstra wrote:
On Thu, 2007-06-07 at 02:17 +0200, Martin Peschke wrote:
Ingo Molnar wrote:
* Martin Peschke <[EMAIL PROTECTED]> wrote:
The output has changed from a terribly wide table to an enormously
long list (just the generic way the statistics code prints data).
Sigh, wh
Dave Hansen wrote:
On Wed, 2007-06-06 at 23:33 +0200, Martin Peschke wrote:
struct statistic_interface {
/* private: */
struct list_head list;
- struct dentry *debugfs_dir;
- struct dentry *data_file;
- struct dentry *def_file
Ingo Molnar wrote:
* Martin Peschke <[EMAIL PROTECTED]> wrote:
- lock_time_inc() vs. statistic_add_util()
please fix the coding style in lib/statistic.c. It's full of:
{
unsigned long long i;
if (value <= stat->u.histogram.range_min)
return
Ingo Molnar wrote:
* Martin Peschke <[EMAIL PROTECTED]> wrote:
The output has changed from a terribly wide table to an enormously
long list (just the generic way the statistics code prints data).
Sigh, why dont you _ask_ before doing such stuff?
A nice diffstat is always worth
e_lock hold_read variance 0.000
clocksource_lock hold_write samples 308908
clocksource_lock hold_write minimum 188
clocksource_lock hold_write average 435.165
clocksource_lock hold_write maximum 2485654
clocksource_lock hold_write variance 86184020.322
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]&g
another_statistic average 0.000
my_entity_2 another_statistic maximum 0
my_entity_2 another_statistic variance 0.000
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
include/linux/statistic.h | 47
lib/statistic.c | 444 --
2
tting rid of seq_file as parameter to the label-function.
Now we just pass a buffer to the label-function.
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
include/linux/statistic.h | 11 ---
lib/statistic.c | 23 +++
2 files changed, 23 insert
Statistics code cleanups. Would not have been needed for the following
changes - just done along the way.
- make function void
- found more handy names for a few identifiers
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
drivers/s390/scsi/zfcp_def.h |4
include/linux/stati
stics library functions have been dropped,
because a couple more users would be nice.)
Ah, and you might need to add CONFIG_STATISTICS to Kconfig of the
architecture of your choice (see s390).
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
-
To unsubscribe from this list: send the line &
Andrew Morton wrote:
On Sat, 2 Jun 2007 19:14:25 +0200
Adrian Bunk <[EMAIL PROTECTED]> wrote:
statistics-infrastructure-make-printk_clock-a-generic-kernel-wide-nsec-resolution.patch
shows why __attribute__((weak)) is harmful because you don't see if a
required non-weak implemtation is missing:
Arjan van de Ven wrote:
On Mon, 2007-05-14 at 10:26 +0200, Martin Peschke wrote:
[EMAIL PROTECTED] timer_stats]# cat data
this patch changes the userspace API though, and breaks PowerTOP :(
Though I think it was unfortunate to add the "old" proc API in 2.6.21.
timer_stats is clear
[EMAIL PROTECTED] wrote:
On Wed, 09 May 2007 01:23:22 PDT, Andrew Morton said:
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21/2.6.21-mm2/
So I'm trying to get the Intel PowerTop stuff working, and discovered that /
proc/timer_stats and/or /proc/tstats has moved to /debug
use Documentation/SubmitChecklist when testing your code ***
See http://www.zip.com.au/~akpm/linux/patches/stuff/added-to-mm.txt to find
out what to do about this
--
Subject: timer_stats slimmed down: using statistics infrastucture
From: Martin
Fengguang Wu wrote:
2007/5/3, Martin Peschke <[EMAIL PROTECTED]>:
This patch implements the bulk of the existing the adaptive readahead
statistics with lib/statistics.c.
The output format differs somewhat as the the statistics code just emits
lists of items and doesn't do ta
looks like:
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
include/linux/statistic.h | 16 +++---
lib/statistic.c | 72 +-
2 files changed, 38 insertions(+), 50 deletions(-)
Index: linux/include/linux/stati
This patch adapts zfcp to the counter changes in lib/statistics.c.
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
zfcp_ccw.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
Index: linux/drivers/s390/scsi/zfcp
cue.none 0 0
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
readahead.c | 170
1 file changed, 34 insertions(+), 136 deletions(-)
Index: linu
zfcp
[RFC] [Patch 3/3] readahead statistics slimmed down
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/maj
ype=sparse entries=1024
data=[6920660.417313] started=[6920660.417313] stopped=[0.00]
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
time/timer_stats.c | 385 -
timer.c|2
2 files changed, 59 insertion
can label entries using their interpretation of a key.
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
include/linux/statistic.h | 132 ++-
lib/statistic.c | 191 +++---
2 files changed, 272 insertions(
down: statistics prereq, labels
[Patch 3/4] timer_stats slimmed down: statistics prereq, keys
[Patch 4/4] timer_stats slimmed down: using statistics infrastucture
Patch set is against 2.6.21-rc6-mm1 and additionaly requires
"[Patch] statistics: fix sorted list" posted this morning.
Signed
Small cleanup
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
include/linux/statistic.h |6 +++---
lib/statistic.c |2 +-
2 files changed, 4 insertions(+), 4 deletions(-)
Index: linux/include/linux/stati
This patch allows clients of the statistics infrastructure to be called
back when an entry of a statistic is printed and if the client might
want to add some kind of label to the output string. A line would look
like this then:
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
i
t
suffer a performance loss.
Fix is to sort the list by hits after entry insertion, i.e. moving
entries added at the tail up the chain, which should not be required
very often (usually just when merging per-cpu lists).
Patch is against 2.6.21-rc6-mm1.
Signed-off-by: Martin Peschke <[EM
We can save some lines of code by using seq_release_private().
Patch is against 2.6.21-rc6-mm1.
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
drivers/md/md.c | 11 +--
fs/proc/proc_misc.c |9 +
kernel/kallsyms.c |9 +
net/sunrpc/cache.c
In case of CPU hotunplug statistics might have leaked some memory
(lists of struct statistic_entry_sparse, about 32 byte each, freed
by statistic_sparse_reset()).
Patch is against 2.6.21-rc6-mm1.
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
statistic.c |1 +
1 file chan
: Martin Peschke <[EMAIL PROTECTED]>
---
statistic.c |6 --
1 file changed, 4 insertions(+), 2 deletions(-)
Index: linux/lib/statistic.c
===
--- linux.orig/lib/statistic.c
+++ linux/lib/statistic.c
@@ -631,11 +631,13 @@
Make sure strsep() in statistic_release_def() finds a terminanal '\0' and
doesn't attempt to access bytes outside the given buffer.
(Patch fixes
statistics-infrastructure-simplify-statistics-debugfs-write-function.patch)
Patch is against 2.6.21-rc6-mm1.
Signed-off-by: Martin
This patch replaces some ugly code implementing debugfs files.
The cleaned up code uses seq_file. File contents remains unchanged.
Patch is against 2.6.21-rc6-mm1.
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
statistic.c
752] [] system_call+0x7e/0x83
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
statistic.c | 111 ++--
1 file changed, 26 insertions(+), 85 deletions(-)
Index: linux/lib/st
Without this fix the the statistics user interface accepted, for example:
echo name=my_stats type=utilisa > definition
while it should only accept:
echo name=my_stats type=utilisation > definition
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
statis
simplifying statistics code by using for_each_substring() and
match_substring() instead of strsep() and match_token()
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
statistic.c | 67 +---
1 file changed, 20 insertions(
() /
match_substring(), e.g. all the stuff in fs/, should not be difficult.
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
include/linux/parser.h |7 +++
lib/parser.c | 45 -
2 files changed, 51 insertions(+), 1 deletion(-)
Index:
simplifying code by using available string duplication function
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
parser.c |8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
Index: linux/lib/pa
The return value of scnprintf() never exceeds @size.
Patch is against 2.6.20-rc6-mm3
Signed-off-by: Martin Peschke <[EMAIL PROTECTED]>
---
vsprintf.c |3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
--- a/lib/vsprintf.c2007-02-02 19:12:09.0 +0100
+++ b/lib/vspr
44 matches
Mail list logo
