linaro-dev

linaro-dev@lists.linaro.org

4332 discussions

by Bernhard Rosenkränzer

== Progress == * Added libatomic to Android toolchain (DEVPLAT-101(?), 30%) * Revamped Android toolchain build to use Bionic from master branch (needed for v8 support), adapt to changes there (DEVPLAT-101(?), 30%) * Cleaned up Android toolchain build scripts while at it (DEVPLAT-101, 20%) * Hunted down fixes for 2 gcc 4.9 bugs that are showstoppers for Android (DEVPLAT-101(?), 10%) * Helped with "Android Toolchain: Member Expectations" document (DEVPLAT-101(?), 10%) == Plans == * Toolchain binary release (DEVPLAT-101) * Android toolchain release (DEVPLAT-101(?)) * Update NDK build scripts to match Android toolchain improvements (DEVPLAT-101(?)) * Investigate any compiler bugs that (might) show up with the new Android toolchain * Finish porting AArch64 Bionic optimizations to ARMv7 (DEVPLAT-104) == Issues == * Not sure the scope of DEVPLAT-101 is wide enough to include all work done last week and upcoming work (esp. NDK), may want to create a few related cards.

11 years, 3 months

[PATCH IDLESTAT] fix execute() function

by Mohammad Merajul Islam Molla

Patch to fix the below two issues - 1. Fix the part where parent waits to fetch child status. In case SIGALRM is received, fetch new status after killing child so that subsequent check for termination status passes. 2. In case of normal termination, just check if program exited normally (checking exit status is not necessary) diff --git a/idlestat.c b/idlestat.c index 787d7b0..179e7a7 100644 --- a/idlestat.c +++ b/idlestat.c @@ -1196,12 +1196,12 @@ static int execute(int argc, char *argv[], char *const envp[], alarm(options->duration); again: if (waitpid(pid, &status, 0) < 0) { - if (errno != EINTR || !sigalrm) - goto again; - kill(pid, SIGTERM); + if (errno == EINTR && sigalrm) + kill(pid, SIGTERM); + goto again; } - if (WIFEXITED(status) && !WEXITSTATUS(status)) { + if (WIFEXITED(status)) { /* * Cancel the timer in case the program * finished before the timeout -- Thanks, -Meraj

11 years, 3 months

[PATCH] enumerate cpu based on different string on ARM

by Sanjay Singh Rawat

bogomips is removed by timer-backed delay loop (commit 9fc2105a) in ARM kernel. Using other suitable string to decide to enumerate each cpu. Signed-off-by: Sanjay Singh Rawat <sanjay.rawat(a)linaro.org> --- src/cpu/cpu.cpp | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/src/cpu/cpu.cpp b/src/cpu/cpu.cpp index b3b6166..105c066 100644 --- a/src/cpu/cpu.cpp +++ b/src/cpu/cpu.cpp @@ -296,6 +296,18 @@ void enumerate_cpus(void) number = -2; } } + /* bogomips is removed in ARM, using CPU revision to enumerate */ + if (strncasecmp(line, "CPU revision\t", 13) == 0) { + if (number == -1) { + /* Not all /proc/cpuinfo include "processor\t". */ + number = 0; + } + if (number >= 0) { + handle_one_cpu(number, vendor, family, model); + set_max_cpu(number); + number = -2; + } + } } -- 1.7.10.4

11 years, 3 months

[PATCH PMQA] nohz: check cpus are in nohz mode

by Sanjay Singh Rawat

check for the timer interrupt raised on cpu for duration of 3 seconds and report the mode Signed-off-by: Sanjay Singh Rawat <sanjay.rawat(a)linaro.org> --- Makefile | 2 ++ nohz/Makefile | 25 +++++++++++++++++++++ nohz/nohz_01.sh | 64 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ nohz/nohz_01.txt | 1 + 4 files changed, 92 insertions(+) create mode 100644 nohz/Makefile create mode 100755 nohz/nohz_01.sh create mode 100644 nohz/nohz_01.txt diff --git a/Makefile b/Makefile index 731619d..896497c 100644 --- a/Makefile +++ b/Makefile @@ -34,6 +34,7 @@ check: @(cd thermal; $(MAKE) check) # @(cd powertop; $(MAKE) check) @(cd cputopology; $(MAKE) check) + @(cd nohz; $(MAKE) check) uncheck: @(cd cpufreq; $(MAKE) uncheck) @@ -41,6 +42,7 @@ uncheck: @(cd cpuidle; $(MAKE) uncheck) # @(cd suspend; $(MAKE) uncheck) @(cd thermal; $(MAKE) uncheck) + @(cd nohz; $(MAKE) uncheck) recheck: uncheck check diff --git a/nohz/Makefile b/nohz/Makefile new file mode 100644 index 0000000..315ff6f --- /dev/null +++ b/nohz/Makefile @@ -0,0 +1,25 @@ +# +# PM-QA validation test suite for the power management on Linux +# +# Copyright (C) 2014, Linaro Limited. +# +# This program 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; either version 2 +# of the License, or (at your option) any later version. +# +# 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; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. +# +# Contributors: +# Sanjay Singh Rawat <sanjay.rawat(a)linaro.org> (LG Electronics) +# - initial API and implementation +# + +include ../Test.mk diff --git a/nohz/nohz_01.sh b/nohz/nohz_01.sh new file mode 100755 index 0000000..2828188 --- /dev/null +++ b/nohz/nohz_01.sh @@ -0,0 +1,64 @@ +#!/bin/bash +# +# PM-QA validation test suite for the power management on Linux +# +# Copyright (C) 2014, Linaro Limited. +# +# This program 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; either version 2 +# of the License, or (at your option) any later version. +# +# 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; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. +# +# Contributors: +# Sanjay Singh Rawat <sanjay.rawat(a)linaro.org> (LG Electronics) +# - initial API and implementation +# + +# URL : https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/P… + +source ../include/functions.sh + +check_cpu_nohz_interrupt() { + local cpu=$1 + local cpuid=${cpu:3} + local ints= + local inte= + local i= + local diff= + + i=$(cat /proc/interrupts | grep time | wc -l) + if [ $i -ne 1 ]; then + echo "warning: multiple entry for timer interrupt :" $i + exit 1 + fi + + if [ "$cpu" == "cpu0" ]; then + return 0 + fi + + cpuid=$((cpuid+2)) + ints=$(cat /proc/interrupts | grep time | awk -v c=$cpuid '{ print $c }') + sleep 3 + inte=$(cat /proc/interrupts | grep time | awk -v c=$cpuid '{ print $c }') + + #echo $cpu "interrupts before" $ints "after" $inte + diff=$((inte-ints)) + + if [ $diff -eq 0 ]; then + echo $cpu "in NOHZ" + else + echo $cpu "not in NOHZ" + fi +} + +for_each_cpu check_cpu_nohz_interrupt +test_status_show diff --git a/nohz/nohz_01.txt b/nohz/nohz_01.txt new file mode 100644 index 0000000..0c547f5 --- /dev/null +++ b/nohz/nohz_01.txt @@ -0,0 +1 @@ +check each cpu is in nohz mode (wait 3 seconds for timer interrupt) -- 1.7.10.4

11 years, 3 months

[idlestat 0/6] More cleanups

by Amit Kucheria

From: Amit Kucheria <amit.kucheria(a)linaro.org> More cleanups to the command-line options to make idlestat easier to use and separate the two modes: tracing and reporting. Regards, Amit Amit Kucheria (6): Use the correct version in the log file Print more details when fopen fails kick the dump option off to use -z and release -m for later -o stands for output file when idlestat can take an input file as well Introduce --trace and --import mode Lots of changes related to usage, bump the version to 0.3 README | 10 ++++----- idlestat.c | 71 +++++++++++++++++++++++++++++++++++++++++++------------------- 2 files changed, 54 insertions(+), 27 deletions(-) -- 1.9.1

11 years, 3 months

Planned maintenance for ci.linaro.org, Friday 6th June

by Milo Casagrande

Hello, this email is to announce that ci.linaro.org will not be available on Friday 6th June due to planned maintenance. We will perform a server upgrade to the latest Ubuntu LTS release. In order to perform a smooth upgrade, please make any update to your jobs before Thursday 5th June at 12:00UTC. Another email will be sent out when the service has been restored. Regards. -- Milo Casagrande Linaro.org <www.linaro.org> │ Open source software for ARM SoCs

11 years, 3 months

idlestat: execute() function problems?

by Mohammad Merajul Islam Molla

Hello, I noticed some odd behaviors with idlestate.c/execute() function. 1. I wrote a sample program hello.c with infinite loop - int main() { while(1); } #sudo ../idlestat/idlestat -o /tmp/myoutput ../temp/hello idlestat never terminates, because there is no alarm set, hello program never terminates. 2. If I change and invoke idlestat with below command - #sudo ../idlestat/idlestat -o /tmp/myoutput -t 5 ../temp/hello idlestat outputs only below line and exits - Total trace buffer: 153896 kB What is happening is the below check fails - if (WIFSIGNALED(status) && WTERMSIG(status) == SIGTERM) { return 0; } here condition is always false. Because the status variable was fetched before killing the process. So status does not contain valid signal number. if (waitpid(pid, &status, 0) < 0) { if (errno != EINTR || !sigalrm) goto again; kill(pid, SIGTERM); } Possible Fix: call waitpid again after kill and fetch new status. 3. The below check makes idlestat dependent on the hello program run - if (WIFEXITED(status) && !WEXITSTATUS(status)) { } If my program is written to return a non-zero value, the second condition will be always false. So with the below program - int main() { return 1; } idlestat outputs only since WEXITSTATUS(status) returns nonzero value - Total trace buffer: 153896 kB Possible Fix: Just check WIFEXITED(status) and ommit the fetching of exit status. In any case, it should not be allowed to run idlestat with "-t 0", since that will results in problem 1 again. -- Thanks, - Meraj

11 years, 3 months

[URGENT] Perf report shared objects showing unknown on ARM

by sneha priya

Hello, There is an issue related to perf which I am facing since 15 days. Hoping that the great minds here will help me to solve this. I have a requirement to make perf tool work on a device having ARM architecture. But, on recording the tracepoint events and then running ./perf report, it shows the shared objects name as [unknown] and Symbols as 00000, whereas for software and hardware events I do not experience this issue.I have cross compiled the perf tool available on mailine and ported it on a device having ARM architecture. Output on ARM based device with kernel 3.4. ./perf record -e kmem:kmalloc cal [ perf record: Woken up 1 times to write data ][ perf record: Captured and wrote 0.007 MB perf.data (~321 samples) ] ./perf report Overhead Command Shared Object Symbol ........................... ................................ ...........................................40.78% cal [unknown] [.]0000000031.6% cal [unknown] [.]00000000 [...] On ubuntu 12.04, system (kernel 3.10) it works perfectly fine. Output on x86 architecture ./perf record -e kmem:kmalloc gcalctool [ perf record: Woken up 3 times to write data ][ perf record: Captured and wrote 0.27 MB perf.data (~845 samples) ] ./perf report --stdio Overhead Command Shared Object Symbol ........................... ................................ ...........................................96.55% cal [kernel.kallsyms] [k] kmem_cache_alloc_trace3.45% cal [kernel.kallsyms] [.]__kmalloc [...] Keenly, awaiting for you help. Thanks. Sneha.

11 years, 3 months

[PATCH POWERDEBUG] gpio: expose free gpios to user

by Sanjay Singh Rawat

- GPIOs in the SoC will be free/used state, this patch make free gpios accessible to the user by exporting them. - bug 1229658 Signed-off-by: Sanjay Singh Rawat <sanjay.rawat(a)linaro.org> --- gpio.c | 27 +++++++++++++++------------ 1 file changed, 15 insertions(+), 12 deletions(-) diff --git a/gpio.c b/gpio.c index 4a42cb7..39480b0 100644 --- a/gpio.c +++ b/gpio.c @@ -320,10 +320,10 @@ static struct display_ops gpio_ops = { .change = gpio_change, }; -void export_gpios(void) +void export_free_gpios(void) { FILE *fgpio, *fgpio_export; - int gpio[256], num = 0; + int i, gpio_max = 0; char *line = NULL; ssize_t read, len; @@ -340,15 +340,18 @@ void export_gpios(void) } /* export the gpios */ - while (read = getline(&line, &len, fgpio) != -1) { - char *str; - - if (strstr(line, "gpio-")) { - str = strtok(line, " "); - sscanf(str, "gpio-%d", &gpio[num]); - fprintf(fgpio_export, "%d", gpio[num]); - num++; - } + while ((read = getline(&line, &len, fgpio)) != -1) { + if (strstr(line, "GPIOs")) + sscanf(line, "%*[^-]-%d%*", &gpio_max); + } + + printf("log: total gpios = %d\n", gpio_max); + for (i = 0 ; i <= gpio_max ; i++) { + char command[50] = ""; + + sprintf(command, "echo %d > /sys/class/gpio/export", i); + if (system(command) < 0) + printf("error: failed to export gpio-%d\n", i); } out: return; @@ -368,7 +371,7 @@ int gpio_init(void) if (access(SYSFS_GPIO, F_OK)) gpio_error = true; /* set the flag */ - export_gpios(); + export_free_gpios(); gpio_tree = tree_load(SYSFS_GPIO, gpio_filter_cb, false); if (!gpio_tree) -- 1.7.10.4

11 years, 3 months

idlestat: possible bug?

by Mohammad Merajul Islam Molla

Hi I was trying to run idlestat and with the below command output is - # sudo ./idlestat -o /tmp/myoutput fopen: No such file or directory I looked into the source and it looks like its not entering in the if condition in main - 1212 /* Acquisition time specified means we will get the traces */ 1213 if (options.duration || args < argc) { 1214 1215 /* Read cpu topology info from sysfs */ 1216 read_sysfs_cpu_topo(); ... 1260 if (idlestat_store(options.filename)) 1261 return -1; } Since duration argument is not given and args is actually equal to argc, if condition is evaluated false. So idlestat_store() is not creating the file "/tmp/myoutput" or storing any info there. As a result, call to idlestat_load is issues above error message. 1264 /* Load the idle states information */ 1265 datas = idlestat_load(options.filename); 1266 if (!datas) 1267 return 1; I think the fix is simply to change the if condition above and make it - 1212 /* Acquisition time specified means we will get the traces */ 1213 if (options.duration || args <= argc) { 1214 } -- Thanks, -Meraj

11 years, 3 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

linaro-dev