What tools and libraries should I use when writing
infrastructure-style projects? Do we have a preferred technology
list? Is there a style guide or standard templates that I can use?
This came about in writing a little web application for tracking the
changes in Linaro GCC vs upstream:
http://ex.seabright.co.nz/patchtrack/
The code is at https://code.launchpad.net/~michaelh1/+junk/patchtrack.
It uses web.py for the app engine, web.py's built-in template engine,
and a hacked up version of the wiki template. I'd love to use tools
that somebody else knows as well though :)
Canonical have a preferred technology list but it's a bit heavy - zope
or django are an overkill for something like this.
Any thoughts?
-- Michael
The weekly report for the Linaro Infrastructure team may be found at:-
https://wiki.linaro.org/Releases/InfrastructureStatus
Summary:
Three LexBuilder <https://wiki.linaro.org/LexBuilder> bugs awaiting
review.
Work ongoing for hosting a public location where we can have live
production/demo of the dashboard.
There is a concern that the changes to LexBuilder are behind schedule.
The first licence request submitted to the TSC was approved. Two
more requests have been submitted to the TSC to cover the test environment.
Progress:
A lot of tasks remain outstanding for the 10.10 beta. This is being
investigated to establish of this is realistic. An update will be
provided in the next report.
This patch gets the available C-states from cpuidle framework (sysfs) and gets the P-states information from the cpufreq (again, from sysfs). Thus it removes the hardcoded values for c and p states, and makes it generic enough to run on all the systems which plugin to cpuidle and cpufreq kernel framework.
Signed-off-by: Amit Arora <amit.arora(a)linaro.org>
---
Makefile | 2 +-
cpufreqstats.c | 3 +-
display.c | 15 ++++++---
generic_cstates.c | 63 ++++++++++++++++++++++++++++++++++++++++
omapcstates.c | 64 -----------------------------------------
powertop.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++-----
powertop.h | 23 +++-----------
7 files changed, 154 insertions(+), 98 deletions(-)
create mode 100644 generic_cstates.c
delete mode 100644 omapcstates.c
diff --git a/Makefile b/Makefile
index 0b41dc4..d9e4883 100644
--- a/Makefile
+++ b/Makefile
@@ -17,7 +17,7 @@ CC?=gcc
OBJS = powertop.o config.o process.o misctips.o bluetooth.o display.o suggestions.o wireless.o cpufreq.o \
sata.o xrandr.o ethernet.o cpufreqstats.o usb.o urbnum.o intelcstates.o wifi-new.o perf.o \
- alsa-power.o ahci-alpm.o dmesg.o devicepm.o omapcstates.o
+ alsa-power.o ahci-alpm.o dmesg.o devicepm.o generic_cstates.o
powertop: $(OBJS) Makefile powertop.h
diff --git a/cpufreqstats.c b/cpufreqstats.c
index b250407..6cd36ee 100644
--- a/cpufreqstats.c
+++ b/cpufreqstats.c
@@ -42,7 +42,7 @@ struct cpufreqdata oldfreqs[16];
struct cpufreqdata delta[16];
-char cpufreqstrings[MAX_NUM_PSTATES+1][80];
+char **cpufreqstrings;
int topfreq = -1;
static void zap(void)
@@ -111,7 +111,6 @@ void do_cpufreq_stats(void)
uint64_t total_time = 0;
memcpy(&oldfreqs, &freqs, sizeof(freqs));
- memset(&cpufreqstrings, 0, sizeof(cpufreqstrings));
sprintf(cpufreqstrings[0], _("P-states (frequencies)\n"));
for (ret = 0; ret<16; ret++)
diff --git a/display.c b/display.c
index cc81efe..de9146b 100644
--- a/display.c
+++ b/display.c
@@ -91,7 +91,12 @@ int maxwidth = 200;
void setup_windows(void)
{
- int yline = MAX_NUM_CSTATES;
+ int yline;
+
+ if (max_num_cstates > max_num_pstates)
+ yline = max_num_cstates;
+ else
+ yline = max_num_pstates;
getmaxyx(stdscr, maxy, maxx);
@@ -152,7 +157,7 @@ void show_title_bar(void)
werase(status_bar_window);
x = 0;
- for (i=0; i < MAX_CSTATE_LINES; i++) {
+ for (i=0; i < max_cstate_lines; i++) {
if (strlen(status_bar_slots[i])==0)
continue;
wattron(status_bar_window, A_REVERSE);
@@ -168,18 +173,18 @@ void show_cstates(void)
int i, count = 0;
werase(cstate_window);
- for (i=0; i < MAX_CSTATE_LINES; i++) {
+ for (i=0; i < max_cstate_lines; i++) {
if (i == topcstate+1)
wattron(cstate_window, A_BOLD);
else
wattroff(cstate_window, A_BOLD);
- if (strlen(cstate_lines[i]) && count <= MAX_CSTATE_LINES) {
+ if (strlen(cstate_lines[i]) && count <= max_cstate_lines) {
print(cstate_window, count, 0, "%s", cstate_lines[i]);
count++;
}
}
- for (i=0; i <= MAX_NUM_PSTATES; i++) {
+ for (i=0; i <= max_num_pstates; i++) {
if (i == topfreq+1)
wattron(cstate_window, A_BOLD);
else
diff --git a/generic_cstates.c b/generic_cstates.c
new file mode 100644
index 0000000..dbee3c1
--- /dev/null
+++ b/generic_cstates.c
@@ -0,0 +1,63 @@
+/*
+ * Copyright 2008, Texas Instruments Incorporated.
+ * Copyright 2010, IBM Corporation
+ *
+ * This file prints the C states supported by any processor.
+ * (Based on intelcstates.c)
+ *
+ * This program file is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ * for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program in a file named COPYING; if not, write to the
+ * Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor,
+ * Boston, MA 02110-1301 USA
+ */
+
+#include <unistd.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdint.h>
+#include <sys/types.h>
+#include <dirent.h>
+#include <ctype.h>
+
+#include "powertop.h"
+
+
+/**
+ * print_generic_cstates() - Prints the list of supported C-states.
+ *
+ * This functions uses standard sysfs interface of the cpuidle framework
+ * to extract the information of the C-states supported by the Linux
+ * kernel.
+ **/
+void print_generic_cstates(void)
+{
+ DIR *dir;
+ struct dirent *entry;
+
+ dir = opendir("/sys/devices/system/cpu/cpu0/cpuidle");
+
+ if (dir) {
+ printf(_("Supported C-states : "));
+
+ while ((entry = readdir(dir))) {
+ if (strlen(entry->d_name) < 3)
+ continue;
+
+ printf("C%s ", entry->d_name);
+ }
+ printf("\n");
+
+ closedir(dir);
+ }
+}
diff --git a/omapcstates.c b/omapcstates.c
deleted file mode 100644
index 9d1da4f..0000000
--- a/omapcstates.c
+++ /dev/null
@@ -1,64 +0,0 @@
-/*
- * Copyright 2008, Texas Instruments Incorporated.
- *
- * This file prints the C states supported by the OMAP processor.
- * (Based on intelcstates.c)
- *
- * This program file is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License as published by the
- * Free Software Foundation; version 2 of the License.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
- * for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program in a file named COPYING; if not, write to the
- * Free Software Foundation, Inc.,
- * 51 Franklin Street, Fifth Floor,
- * Boston, MA 02110-1301 USA
- */
-
-#include <unistd.h>
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-#include <stdint.h>
-#include <sys/types.h>
-#include <dirent.h>
-#include <ctype.h>
-
-#include "powertop.h"
-
-
-#if defined(OMAP3)
-/**
- * print_omap3_cstates() - Prints the list of supported C-states.
- *
- * This functions uses standard sysfs interface of the cpuidle framework
- * to extract the information of the C-states supported by the Linux
- * kernel.
- **/
-void print_omap3_cstates(void)
-{
- DIR *dir;
- struct dirent *entry;
-
- dir = opendir("/sys/devices/system/cpu/cpu0/cpuidle");
-
- if (dir) {
- printf(_("Supported C-states : "));
-
- while ((entry = readdir(dir))) {
- if (strlen(entry->d_name) < 3)
- continue;
-
- printf("C%s ", entry->d_name);
- }
- printf("\n");
-
- closedir(dir);
- }
-}
-#endif
diff --git a/powertop.c b/powertop.c
index 7a32623..6fd3efe 100644
--- a/powertop.c
+++ b/powertop.c
@@ -40,6 +40,7 @@
#include "powertop.h"
+extern void print_generic_cstates(void);
uint64_t start_usage[8], start_duration[8];
uint64_t last_usage[8], last_duration[8];
@@ -85,6 +86,70 @@ time_t prev_bat_time = 0;
double displaytime = 0.0;
+char **cstate_lines;
+
+void init_numstates(void)
+{
+ DIR *dir;
+ FILE *fp;
+ struct dirent *entry;
+ int i, count = 0;
+ char line[1024];
+
+ dir = opendir("/sys/devices/system/cpu/cpu0/cpuidle");
+ if (dir) {
+ while ((entry = readdir(dir))) {
+ if (strlen(entry->d_name) < 3)
+ continue;
+ count++;
+ }
+ closedir(dir);
+ }
+ max_num_cstates = count;
+ count = 0;
+
+ fp = fopen("/sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state",
+ "r");
+ if (fp) {
+ while (fgets(line, sizeof(line), fp)!=NULL)
+ count++;
+ fclose(fp);
+ }
+ max_num_pstates = count;
+
+ if (max_num_cstates > max_num_pstates)
+ max_cstate_lines = max_num_cstates + 3;
+ else
+ max_cstate_lines = max_num_pstates + 3;
+
+ /* Now, allocate memory */
+
+ cpufreqstrings = (char **)calloc((max_num_pstates+1), sizeof(char *));
+ assert(cpufreqstrings!=0);
+ for (i=0; i<max_num_pstates+1; i++) {
+ cpufreqstrings[i] = (char *)calloc(80, sizeof(char));
+ assert(cpufreqstrings[i]!=0);
+ }
+
+ cstate_lines = (char **)calloc(max_cstate_lines, sizeof(char *));
+ assert(cstate_lines!=0);
+ for (i=0; i<max_cstate_lines; i++) {
+ cstate_lines[i] = (char *)calloc(200, sizeof(char));
+ assert(cstate_lines[i]!=0);
+ }
+
+}
+
+void free_numstates(void)
+{
+ int i;
+
+ for (i=0; i<max_num_pstates+1; i++)
+ free(cpufreqstrings[i]);
+ free(cpufreqstrings);
+ cpufreqstrings = NULL;
+}
+
void push_line(char *string, int count)
{
int i;
@@ -824,8 +889,6 @@ void print_battery_sysfs(void)
show_acpi_power_line(rate, cap, prev_bat_cap - cap, time(NULL) - prev_bat_time);
}
-char cstate_lines[MAX_CSTATE_LINES][200];
-
void usage()
{
printf(_("Usage: powertop [OPTION...]\n"));
@@ -902,6 +965,8 @@ int main(int argc, char **argv)
memcpy(last_usage, start_usage, sizeof(last_usage));
memcpy(last_duration, start_duration, sizeof(last_duration));
+ init_numstates();
+
do_proc_irq();
do_proc_irq();
do_cpufreq_stats();
@@ -924,8 +989,8 @@ int main(int argc, char **argv)
printf("\n\n");
#if defined (__I386__)
print_intel_cstates();
-#elif defined (OMAP3)
- print_omap3_cstates();
+#else
+ print_generic_cstates();
#endif
stop_timerstats();
@@ -963,7 +1028,7 @@ int main(int argc, char **argv)
totalticks = 0;
totalevents = 0;
- for (i = 0; i < MAX_NUM_CSTATES; i++)
+ for (i = 0; i < max_num_cstates; i++)
if (cur_usage[i]) {
totalticks += cur_duration[i] - last_duration[i];
totalevents += cur_usage[i] - last_usage[i];
@@ -978,8 +1043,7 @@ int main(int argc, char **argv)
show_title_bar();
}
- memset(&cstate_lines, 0, sizeof(cstate_lines));
- topcstate = -(MAX_NUM_CSTATES);
+ topcstate = -(max_num_cstates);
if (totalevents == 0 && maxcstate <= 1) {
sprintf(cstate_lines[5],_("< Detailed C-state information is not available.>\n"));
} else {
@@ -993,7 +1057,7 @@ int main(int argc, char **argv)
sprintf(cstate_lines[1], _("C0 (cpu running) (%4.1f%%)\n"), percentage);
if (percentage > 50)
topcstate = 0;
- for (i = 0; i < MAX_NUM_CSTATES; i++)
+ for (i = 0; i < max_num_cstates; i++)
if (cur_usage[i]) {
sleept = (cur_duration[i] - last_duration[i]) / (cur_usage[i] - last_usage[i]
+ 0.1) / FREQ;
@@ -1300,5 +1364,7 @@ int main(int argc, char **argv)
}
end_data_dirty_capture();
+
+ free_numstates();
return 0;
}
diff --git a/powertop.h b/powertop.h
index 21f5538..4017cfe 100644
--- a/powertop.h
+++ b/powertop.h
@@ -30,19 +30,9 @@
#define VERSION "1.12"
-#if defined(__i386__)
-#define MAX_NUM_CSTATES 4
-#define MAX_NUM_PSTATES 5
-
-#elif defined(OMAP3)
-#define MAX_NUM_CSTATES 7
-#define MAX_NUM_PSTATES 5
-
-#else
-#error "No valid architecture is defined."
-#endif
-
-#define MAX_CSTATE_LINES (MAX_NUM_CSTATES + 3)
+int max_num_cstates;
+int max_num_pstates;
+int max_cstate_lines;
struct line {
char *string;
@@ -83,11 +73,8 @@ void usb_activity_hint(void);
void devicepm_activity_hint(void);
-
-
-
-extern char cstate_lines[MAX_CSTATE_LINES][200];
-extern char cpufreqstrings[MAX_NUM_PSTATES+1][80];
+extern char **cpufreqstrings;
+extern char **cstate_lines;
extern int topcstate;
extern int topfreq;
--
1.7.0.4
Currently the PowerTOP tool is x86 centric and does not work on other architectures, unless some changes are made.
This patchset removes the hardcoded values for the P and C states. It does that in two steps
1) Implements the OMAP3 support in PowerTOP.
2) Makes the code generic enough for all boards
I have tested this only on my desktop. I am yet to receive an ARM based board, and hence couldn't test it there. So, any kind of testing and/or feedback are most welcome! Thanks!
Amit Arora (2):
Add OMAP3 support to PowerTOP
Remove hardcoded c and p states
Makefile | 2 +-
cpufreqstats.c | 3 +-
display.c | 23 +++++++++-----
generic_cstates.c | 63 ++++++++++++++++++++++++++++++++++++++++
powertop.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++++----
powertop.h | 11 ++++---
6 files changed, 162 insertions(+), 22 deletions(-)
create mode 100644 generic_cstates.c
--
Regards,
Amit Arora
Hi,
These are some very rough instructions for getting Maverick running on
an IGEPv2 board. Maybe they should go on the wiki somewhere, but I
wasn't sure where.
Running Maverick over NFS on an IGEPv2 board
============================================
I have the Ubuntu 9.04 "demo" image on a microSD card, and wanted to try running Maverick. To reproduce this recipe, you'll need the following:
* microSD card reader
* An installed Maverick rootfs (mine was created with qemu-debootstrap) on a system running an NFS server.
* An IGEPv2 serial cable might be handy too.
Part 1, new kernel with existing Jaunty installation
----------------------------------------------------
The kernel which comes with the demo image, 2.6.28.10 (with IGEPv2 patches), has a few problems which prevent its use with Maverick. You need to use a newer version.
Grab a precompiled kernel (I used uImage-2.6.33.5-0-igep0020.bin) from:
http://labs.igep.es/index.php/The_Linux_kernel
Now we need to make a ".ini" file which will boot this kernel. To use the existing (Jaunty) distribution on the SD card, use a config file like this:
--- cut ---
mmc init 0
echo "===== Setup boot settings ====="
setenv bootargs-base 'mem=512M console=ttyS2,115200n8 console=tty0 omapfb.mode=dvi:1024x768MR-16@60'
setenv mmc-bootargs 'setenv bootargs ${bootargs-base} root=/dev/mmcblk0p2 rw rootwait '
echo "===== Load uImage from mmc ====="
run mmc-bootargs
fatls mmc 0:1
fatload mmc 0 80200000 uImage-2.6.33.5-0-igep0020.bin
bootm 80200000
--- cut ---
Call this file "setup-ini.source". Now install the package "uboot-mkimage" on your host system and run:
mkimage -A arm -O linux -T script -C none -a 0 -e 0 -n 'Boot setup script' -d \
setup-ini.source setup.ini
Copy setup.ini to the boot partition of the microSD card (backing up the original).
Now: you may find this is not sufficient to make the IGEPv2 boot. The problem is apparently bugs in the FAT support in the supplied version of uboot. This can be seen over the serial line in the output of "fatls": your kernel image should appear in the list.
If it does not, you might be able to work around it by doing the following:
* Copy all files off the boot partition.
* Reformat the boot partition (with the microSD card connected via a card reader to a Linux host) with:
mkdosfs /dev/sdX1
(note, adding "-F 32" did *not* work).
* Copy all files back to the boot partition.
* Don't forget to unmount!
Now, reinsert the card and Jaunty should boot with the new kernel (check with "uname -a").
Part 2, new kernel with Maverick over NFS root
----------------------------------------------
To boot with an NFS root, you will need a new setup.ini. Use the following source file:
== cut ==
mmc init 0
echo "===== Setup boot settings ====="
setenv bootargs-base 'mem=512M console=ttyS2,115200n8 console=tty0 omapfb.mode=dvi:1024x768MR-16@60'
setenv mmc-bootargs 'setenv bootargs ${bootargs-base} ip=dhcp root=/dev/nfs nfsroot=192.168.1.64:/mnt/bitbucket/jules/linaro/m-armel rw rootwait'
echo "===== Load uImage from mmc ====="
run mmc-bootargs
fatls mmc 0:1
fatload mmc 0 80200000 uImage-2.6.33.5-0-igep0020.bin
bootm 80200000
== cut ==
Substitute your NFS server for the "nfsroot" IP address, and the path on that server for the subsequent path. Use the instructions above to turn this into setup.ini, and copy to the microSD card (noting that the issue about FAT bugs still applies).
On the server, use a line like the following in /etc/exports (making local changes as necessary):
/mnt/bitbucket/jules/linaro/m-armel *.config(rw,no_root_squash,no_subtree_check,sync)
Note that "portmap" should be installed rather than "rpcbind", if your distribution offers a choice between the two (at least my board wouldn't boot, at least with the default configuration of the latter). Use "nfs-kernel-server" rather than "nfs-user-server", again if your distro offers the choice.
It's probably possible to use tftp/bootp to load a kernel instead: I
haven't experimented with that so far.
Dnia czwartek, 29 lipca 2010 o 13:57:21 Amit Arora napisał(a):
> I have tested this only on my desktop. I am yet to receive an ARM based
> board, and hence couldn't test it there. So, any kind of testing and/or
> feedback are most welcome! Thanks!
I fetched 9f4efd1c7cc0bc9a9078fa6b83a98763a7637e2f revision, applied both
patches and built it for my desktop. Segfaulted on first run.
Desktop is Core2Quad q6600, powertop without your changes works. C-states are
not available, frequencies are listed.
Regards,
--
JID: hrw(a)jabber.org
Website: http://marcin.juszkiewicz.com.pl/
LinkedIn: http://www.linkedin.com/in/marcinjuszkiewicz
Tim,
Based on your suggestions last week I have prepared a new linux-linaro
tree. The source package name is now linux-linaro. The debian.linaro
changelog only has high level changes not all the git commits. I
removed all the cruft from the debian.linaro subtree. The url is:
git://git.linaro.org/jcrigby/linux-linaro.git
The linaro and master branches are currently identical. The latest
tag is Linaro-2.6.35-1000.2. A source package based on this is in a
ppa building at this very moment.
Thanks
John
The Agenda and the Minutes for this meeting can also be found at
https://wiki.linaro.org/WorkingGroups/PowerManagement/Meetings/2010-07-28
Attendees : Amit Arora, Amit Kucheria, Bobby Bhatacharia, Srinivas
Kalaga, Vishwanath Sripathy
Minutes:
* Review action items from last meeting
o ACTION: Bobby to investigate android contraints (e.g. display keep-alive)
Will target it for the next week.
o ACTION: Srinivas to provide details of where he believes
userspace - kernel interaction is required.
pm_qos is low prio .. so this may take a few weeks for this action
This may take a few weeks for this action. Currently at low
priority (since pm_qos work is of low prio)
o ACTION: Srinivas to add ftrace item to task list and send patches to linaro
Srinivas has been working on this in last few days. We should see
something soon on this.
o ACTION: Amit K. to inform ARM about multi-core availability
We are not getting any multi-core boards. Vishwa should have one.
Now, the question is - will one board be enough ?
Bobby to see possibility of getting more boards for Linaro. Not
easy, but will consider the request open.
o ACTION: ARM to discuss giving out powertop tool
ARM doesn't have a PowerTOP tool, so to speak, to give out.
Though they have an Eclipse based tool, which is very much an ongoing
work. Can work on making it available. But, no ETA as of now.
o ACTION: TI to elaborate on requirement of cpuidle instrumentation task
Vishwa : The currently "known" c-states sleep time and latency
information on OMAP boards are NOT optimal. There is a requirement for
some kind of instrumentation to measure the s/w and the h/w latencies.
Later is difficult part. And once we can get the optimal latencies, we
will have to work on optimizing them.
NEWACTION : Vishwa to create a blue print for this. It should
mention the codepaths to be measured (eg. wakeup latencies, idle loop
to idle loop latencies for various c-states) and other details.
o ACTION: ARM to share internal instrumentation flow
(BAB: we might also align with Linaro on workload discussions)
Bobby to make this available. But, it might take a couple of months.
* HW availability
o Amit Arora is following up to expedite the process to acquire the
development boards. May take some more time.
o Vishwa has access to couple of OMAP3/4 SDP boards. Amit Kucheria
mentioned that it may be a good idea to get the publically available
boards, like the Beagleboards or the PANDA blades. Vishwa to find out
about this.
o Amit Kucheria has a Freescale i.mx51 babbage board and a beagleboard
* Git trees on git.linaro.org
o We will have to host a couple of git trees in git.linaro.org
o To start with, we may need two trees for the tools - for powertop
and powerdebug (a new tool).
o Need access to the host system and some process needs to be followed.
o Once we start hosting the trees, we will have to send a mail to
linaro-dev mailing list, requesting everyone to try out the code.
o When we have patches for kernel, we may have to host a separate
kernel git tree for Power Management, which should get merged with the
linaro kernel git tree.
* gobby / ubuntu
o Amit Kucheria recommended everyone to start using this tool,
which is a cross-platform collaboration tool and will make things like
taking minutes much easier (everyone will be able to contribute to it)
* Blueprint status
o Tools blueprint : Amit Arora gave the update on the work he has
been doing on powertop and powerdebug tools. He has a few patches
ready, which he plans to host in a git tree. For the time being, we
plan to host this code as tarballs at some location, which will help
others to test it.
o Master blueprint : Amit Kucheria had split the master blue print
into "infrastructure" and "tools" blueprints. More such blueprints
will need to be created as we go forward.
o There are some known issues with Ondemand governor, which Vishwa
and AmitK will take it to the the community (LKML/linux-pm and make
sure to copy right people). Will need to share the constraints and ask
for feedback on how its best to implement some of the things in the
current framework.
* Others
o Linaro's first released will be based on 2.6.35. Hence, all our
work also should be inline to this release.
ACTION ITEMS:
Carried Forward from past :
-------------------------
* CLOSED :
o ACTION: TI to elaborate on requirement of cpuidle instrumentation task
This has given birth to a new ACTION item, discussed below.
* ACTIVE :
o ACTION: Bobby to investigate android contraints (e.g. display keep-alive)
o ACTION: Srinivas to provide details of where he believes
userspace - kernel interaction is required.
o ACTION: Srinivas to add ftrace item to task list and send patches to linaro
o ACTION: Amit K. to inform ARM about multi-core availability
o ACTION: ARM to discuss giving out powertop tool
* DORMANT :
o ACTION: ARM to share internal instrumentation flow
(BAB: we might also align with Linaro on workload discussions)
Might take couple of months.
New Items:
---------
o ACTION: Vishwa to create a new blue print for the instrumentation
of cpuidle related work
o ACTION: Amit Arora to create tarballs of his current work and
share it internally for testing
o ACTION: Amit Kucheria and Vishwa to get inputs from community on
the issues related to Ondemand governor.
Thanks!
Amit Arora
