linaro-dev February 2013

linaro-dev@lists.linaro.org

85 participants
82 discussions

[ANNOUNCE] linux-linaro kernel schedule for 13.02

by Andrey Konovalov

Greetings, The current 13.02 schedule has been published on wiki: https://wiki.linaro.org/Platform/DevPlatform/LinuxLinaroKernelSchedule llct rebuild is planned later today. The 13.02 release would be v3.8-rc8 based (tentative). Please make sure that requests to add new topics (with the git branch to pull from) have been sent to me, or will be sent as soon as the topic is ready. I would appreciate a "heads up" like information on the new topics which are more complex than a couple patches on top of mainline. For the moment I am aware of the kvm topic only. Anything else? Thanks, Andrey

11 years, 7 months

Unable to install svn on Ubuntu 12.04 (64-bit)

by Amar Shankar

Hi, We have installed Ubuntu 12.04 (Precise Pangolin) 64-bit. But unable to install svn. rajkiran@rajkiran-OptiPlex-780:~$ sudo apt-get install subversion Reading package lists... Error! W: Duplicate sources.list entry cdrom://Ubuntu 12.04.1 LTS _Precise Pangolin_ - Release amd64 (20120823.1)/ precise/main i386 Packages (/var/lib/apt/lists/Ubuntu%2012.04.1%20LTS%20%5fPrecise%20Pangolin%5f%20-%20Release%20amd64%20(20120823.1)_dists_precise_main_binary-i386_Packages) W: Duplicate sources.list entry cdrom://Ubuntu 12.04.1 LTS _Precise Pangolin_ - Release amd64 (20120823.1)/ precise/restricted i386 Packages (/var/lib/apt/lists/Ubuntu%2012.04.1%20LTS%20%5fPrecise%20Pangolin%5f%20-%20Release%20amd64%20(20120823.1)_dists_precise_restricted_binary-i386_Packages) E: Encountered a section with no Package: header E: Problem with MergeList /var/lib/apt/lists/extras.ubuntu.com_ubuntu_dists_precise_main_binary-amd64_Packages E: The package lists or status file could not be parsed or opened. Could you please help us resolve the issue? Regards, Amar ________________________________ From: Dave Pigott [dave.pigott(a)linaro.org] Sent: 19 February, 2013 4:59 PM To: Amar Shankar Cc: linaro-dev(a)lists.linaro.org; Prasad Raju; Rajkiran Mulugu Subject: Re: Failure to generate beagle_sd.img using linaro-media-create As James said, Anything from the LTS (12.04 - Precise) onwards. Dave On 19 Feb 2013, at 10:02, Amar Shankar <Amar.Shankar(a)infotech-enterprises.com<mailto:Amar.Shankar@infotech-enterprises.com>> wrote: No ________________________________ DISCLAIMER: This email may contain confidential information and is intended only for the use of the specific individual(s) to which it is addressed. If you are not the intended recipient of this email, you are hereby notified that any unauthorized use, dissemination or copying of this email or the information contained in it or attached to it is strictly prohibited. If you received this message in error, please immediately notify the sender at Infotech and delete the original message.

11 years, 7 months

Test Result Summary of Linaro 13.02 Release for Linux Linaro ubuntu Quantal.

by Botao Sun

Linaro 13.02 Release (Calendar Week 9): Here is test result summary for Linux Linaro ubuntu Quantal image on following boards: 1) ARM Versatile Express A9; 2) Samsung Arndale; 3) TI Panda 4430; 4) TI Panda 4460; 5) ST Ericsson Snowball. Synopsis: Kernel version on all boards has been upgraded to "3.8.0-1"; TI Panda 4460 booting failed. 1. ARM Versatile Express A9 + Linux Linaro Quantal (Column M): https://docs.google.com/a/linaro.org/spreadsheet/ccc?key=0AroPySpr4FnEdFNmV… It keeps exactly same status since calendar week 50 last year: only "Halt" & "Device Tree" test failed, all other features work well. 2. Samsung Arndale + Linux Linaro Quantal (Column G): https://docs.google.com/a/linaro.org/spreadsheet/ccc?key=0AgB-fT5LL31CdGZJS… It keeps same status as last week: system hangs during the Power Management test, HDMI display is unavailable. All other features work well. 3. TI Panda 4430 + Linux Linaro Quantal (Column M): https://docs.google.com/a/linaro.org/spreadsheet/ccc?key=0AroPySpr4FnEdEwwZ… USB Host, Ethernet and Power Management back to work, WiFi, HDMI and DVI-D are still unavailable. All other features work well. 4. TI Panda 4460 + Linux Linaro Quantal (Column M): https://docs.google.com/a/linaro.org/spreadsheet/ccc?key=0AroPySpr4FnEdEwwZ… System booting failed since last week, all test cases are blocked. No release for this board in this monthly cycle. 5. ST Ericsson Snowball + Linux Linaro Quantal (Column M): https://docs.google.com/a/linaro.org/spreadsheet/ccc?key=0AroPySpr4FnEdFJ4X… It keeps same status as last week: Power Management, Reboot and HDMI display don't work, also many error logs show about SD MMC I/O. All other features work well. For the previous week test summary (Calendar Week 8), please refer to attachment. Thank you. Best Regards Botao Sun

11 years, 7 months

[PATCH v8 0/2] Add display-timing node parsing to exynos drm fimd

by Vikas Sajjan

Add display-timing node parsing to drm fimd and depends on the display helper patchset at http://lists.freedesktop.org/archives/dri-devel/2013-January/033998.html It also adds pinctrl support for drm fimd. changes since v7: - addressed comments from Joonyoung Shim <jy0922.shim(a)samsung.com> to remove a unnecessary variable. changes since v6: addressed comments from Inki Dae <inki.dae(a)samsung.com> to separated out the pinctrl functionality and made a separate patch. changes since v5: - addressed comments from Inki Dae <inki.dae(a)samsung.com>, to remove the allocation of 'fbmode' and replaced '-1'in "of_get_fb_videomode(dev->of_node, fbmode, -1)" with OF_USE_NATIVE_MODE. changes since v4: - addressed comments from Paul Menzel <paulepanter(a)users.sourceforge.net>, to modify the commit message changes since v3: - addressed comments from Sean Paul <seanpaul(a)chromium.org>, to modify the return values and print messages. changes since v2: - moved 'devm_pinctrl_get_select_default' function call under 'if (pdev->dev.of_node)', this makes NON-DT code unchanged. (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) changes since v1: - addressed comments from Sean Paul <seanpaul(a)chromium.org> Vikas Sajjan (2): video: drm: exynos: Add display-timing node parsing using video helper function video: drm: exynos: Add pinctrl support to fimd drivers/gpu/drm/exynos/exynos_drm_fimd.c | 34 ++++++++++++++++++++++++++---- 1 file changed, 30 insertions(+), 4 deletions(-) -- 1.7.9.5

11 years, 7 months

[PATCH v4] sched: fix init NOHZ_IDLE flag

by Vincent Guittot

On my smp platform which is made of 5 cores in 2 clusters, I have the nr_busy_cpu field of sched_group_power struct that is not null when the platform is fully idle. The root cause seems to be: During the boot sequence, some CPUs reach the idle loop and set their NOHZ_IDLE flag while waiting for others CPUs to boot. But the nr_busy_cpus field is initialized later with the assumption that all CPUs are in the busy state whereas some CPUs have already set their NOHZ_IDLE flag. During the initialization of the sched_domain, we set the NOHZ_IDLE flag when nr_busy_cpus is initialized to 0 in order to have a coherent configuration. If a CPU enters idle and call set_cpu_sd_state_idle during the build of the new sched_domain it will not corrupt the initial state set_cpu_sd_state_busy is modified and clears the NOHZ_IDLE only if a non NULL sched_domain is attached to the CPU (which is the case during the rebuild) Change since V3; - NOHZ flag is not cleared if a NULL domain is attached to the CPU - Remove patch 2/2 which becomes useless with latest modifications Change since V2: - change the initialization to idle state instead of busy state so a CPU that enters idle during the build of the sched_domain will not corrupt the initialization state Change since V1: - remove the patch for SCHED softirq on an idle core use case as it was a side effect of the other use cases. Signed-off-by: Vincent Guittot <vincent.guittot(a)linaro.org> --- kernel/sched/core.c | 4 +++- kernel/sched/fair.c | 9 +++++++-- 2 files changed, 10 insertions(+), 3 deletions(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 26058d0..c730a4e 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -5884,7 +5884,9 @@ static void init_sched_groups_power(int cpu, struct sched_domain *sd) return; update_group_power(sd, cpu); - atomic_set(&sg->sgp->nr_busy_cpus, sg->group_weight); + atomic_set(&sg->sgp->nr_busy_cpus, 0); + set_bit(NOHZ_IDLE, nohz_flags(cpu)); + } int __weak arch_sd_sibling_asym_packing(void) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 81fa536..2701a92 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5403,15 +5403,20 @@ static inline void set_cpu_sd_state_busy(void) { struct sched_domain *sd; int cpu = smp_processor_id(); + int clear = 0; if (!test_bit(NOHZ_IDLE, nohz_flags(cpu))) return; - clear_bit(NOHZ_IDLE, nohz_flags(cpu)); rcu_read_lock(); - for_each_domain(cpu, sd) + for_each_domain(cpu, sd) { atomic_inc(&sd->groups->sgp->nr_busy_cpus); + clear = 1; + } rcu_read_unlock(); + + if (likely(clear)) + clear_bit(NOHZ_IDLE, nohz_flags(cpu)); } void set_cpu_sd_state_idle(void) -- 1.7.9.5

11 years, 7 months

[RFC/PATCH 0/5] context_tracking: prerequisites for ARM support

by Kevin Hilman

Frederic, Here is a first pass at some changes needed in order to add ARM support. I have another series coming that instruments the syscalls, exceptions etc. on ARM also, but this one is just to get some prerequisites out for broader discussion. This series applies on top of your 3.8-rc6-nohz4 branch, and is also available here: git://git.kernel.org/pub/scm/linux/kernel/git/khilman/linux.git wip/arm-nohz/prereq I'm aware of the >80 column warnings from checkpatch on the atomic64 conversion patch, but wanted to get some feedback on this approach before fixing all those up. Maybe some simpler wrappers in kernel_stat.h would make this a bit cleaner? Kevin Kevin Hilman (5): context tracking: conditionalize guest support based on CONFIG_KVM kernel_cpustat: convert to atomic 64-bit accessors virt CPU accounting: Kconfig: drop 64-bit requirment cputime: use do_div() for nsec resolution conversion helpers ARM: Kconfig: allow virt CPU accounting arch/arm/Kconfig | 1 + fs/proc/stat.c | 36 +++++++++++++------------- fs/proc/uptime.c | 2 +- include/asm-generic/cputime_nsecs.h | 51 +++++++++++++++++++++++++++---------- include/linux/kernel_stat.h | 2 +- init/Kconfig | 2 +- kernel/context_tracking.c | 4 +++ kernel/sched/core.c | 10 ++++---- kernel/sched/cputime.c | 24 ++++++++--------- 9 files changed, 80 insertions(+), 52 deletions(-) -- 1.8.1.2

11 years, 7 months

big.LITTLE MP Scheduler test summary-(week-9)-13.02

by Naresh Kamboju

Hi, Release test report (week-9) 13.02: Here is test result summary for big.LITTLE MP Scheduler test on TC2 platform with Android image sched_tests.git No of Test Cases Tests Run Tests Pass Tests Fail Absolute pass rate (%) Failure Analysis/Comments Regression 20 19 19 0 100 - mpbasicsuite 10 9 9 1 90 - mpcoresuite 4 4 4 0 75 - mpextendedsuite 6 6 4 2 66.7 - mploadbalance 2 2 2 0 100 - Android Build: https://android-build.linaro.org/builds/~linaro-android/vexpress-linaro-mp-… Linux kernel: Linux version 3.8.0-00928-g7d866cd (jenkins-build@ip-10-10-66-148 ) (gcc version 4.7.3 20130205 (prerelease) (Linaro GCC 4.7-2013.02) ) #1 SMP Sat Feb 23 11:31:29 UTC 2013 Results spread sheet: https://docs.google.com/a/linaro.org/spreadsheet/ccc?key=0Ai9ggMs8bsGJdER3a… summary: No regression or new bugs found in this test cycle. Results are same as week-8. Best regards Naresh Kamboju

11 years, 7 months

[PATCH v3 0/3] Add ST-Ericsson AB8500 HWMON driver

by Hongbo Zhang

Guenter, Please check this v3 patches, thanks. Anton Vorontsov and David Woodhouse, Please have a look at this 1/3 and 2/3 patches, thanks. v2 -> v3 changes: - Add interface for converting voltage to temperature - Remove temp5 sensor since we cannot offer temperature read interface of it - Update hyst to use absolute temperature instead of a difference - Add the 3/3 patch v1 -> v2 changes: - Add Documentation/hwmon/abx500 and Documentation/hwmon/abx500 - Make devices which cannot report milli-Celsius invisible - Add temp5_crit interface - Re-work the old find_active_thresholds() to threshold_updated() - Reset updated_min_alarm and updated_max_alarm at the end of each loop - Update the hyst mechamisn to make it works as real hyst - Remove non-stand attributes - Re-order the operations sequence inside probe and remove functions - Update all the lock usages to eliminate race conditions - Make attibutes index starts from 0 also changes: - Since the old [1/2] "ARM: ux500: rename ab8500 to abx500 for hwmon driver" has been merged by Samuel, so won't send it again. - Add another new patch "ab8500_btemp: export two symblols" as [2/2] of this patch set. Hongbo Zhang (3): ab8500_btemp: make get_temp interface public ab8500: make res_to_temp tables public hwmon: add ST-Ericsson ABX500 hwmon driver Documentation/hwmon/ab8500 | 22 ++ Documentation/hwmon/abx500 | 26 ++ drivers/hwmon/Kconfig | 13 + drivers/hwmon/Makefile | 1 + drivers/hwmon/ab8500.c | 178 +++++++++++++ drivers/hwmon/abx500.c | 501 +++++++++++++++++++++++++++++++++++ drivers/hwmon/abx500.h | 87 ++++++ drivers/power/ab8500_bmdata.c | 8 +- drivers/power/ab8500_btemp.c | 2 +- include/linux/mfd/abx500/ab8500-bm.h | 5 + 10 files changed, 840 insertions(+), 3 deletions(-) create mode 100644 Documentation/hwmon/ab8500 create mode 100644 Documentation/hwmon/abx500 create mode 100644 drivers/hwmon/ab8500.c create mode 100644 drivers/hwmon/abx500.c create mode 100644 drivers/hwmon/abx500.h -- 1.8.0

11 years, 7 months

builds unstable - IOERROR: BUILD-INFO.txt is not present in build.

by Fathi Boudra

Hi, Several builds became "unstable" (in jenkins term) without changes in the jobs: IOERROR: BUILD-INFO.txt is not present in build. Obviously, it seems BUILD-INFO.txt is now enforced and some jobs are still using EULA.txt mechanism. While I don't have any issues with (en)forcing users to use BUILD-INFO.txt, I would have expected some notifications to allow job maintainers to adjust the builds as appropriate. Once again, it isn't the kind of changes to deploy the release week. Cheers, -- Fathi Boudra Release Manager | Developer Platform Tech Lead Linaro.org | Open source software for ARM SoCs

11 years, 7 months

[PATCH] pmqa: cpuidle: add test to check cpuidle statistics

by Sanjay Singh Rawat

From: Daniel Lezcano <daniel.lezcano(a)linaro.org> Add functionality to record the current runtime cpuidle statistics and show the same. Signed-off-by: Sanjay Singh Rawat <sanjay.rawat(a)linaro.com> --- README | 6 + Test.mk | 6 + cpuidle/cpuidle_04.sh | 41 ++++ cpuidle/cpuidle_04.txt | 1 + cpuidle/cpuidle_stats.c | 472 +++++++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 526 insertions(+) create mode 100755 cpuidle/cpuidle_04.sh create mode 100644 cpuidle/cpuidle_04.txt create mode 100644 cpuidle/cpuidle_stats.c diff --git a/README b/README index a22a3c8..41c5f41 100644 --- a/README +++ b/README @@ -10,3 +10,9 @@ If you want to run a subset of the tests, do: make -C sched_mc check make -C cpufreq check + +For running cpuidle-stats test, which is 4th subtest in cpuidle; run: + + make -C cpuidle check TEST=4 + +Note: Any other value for TEST will not run any other subtest diff --git a/Test.mk b/Test.mk index 671bbf5..19d7346 100644 --- a/Test.mk +++ b/Test.mk @@ -36,7 +36,13 @@ SANITY_STATUS:= $(shell if test $(SNT) && test -f $(SNT); then \ echo 1; fi; else echo 1; fi) ifeq "$(SANITY_STATUS)" "1" + +ifeq "$(TEST)" "4" +TST=cpuidle_04.sh +run_tests: uncheck $(EXEC) $(LOG) +else run_tests: uncheck $(EXEC) $(LOG) +endif %.log: %.sh @echo "###" diff --git a/cpuidle/cpuidle_04.sh b/cpuidle/cpuidle_04.sh new file mode 100755 index 0000000..42d6de5 --- /dev/null +++ b/cpuidle/cpuidle_04.sh @@ -0,0 +1,41 @@ +#!/bin/bash +# +# PM-QA validation test suite for the power management on Linux +# +# Copyright (C) 2011, 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: +# Daniel Lezcano <daniel.lezcano(a)linaro.org> (IBM Corporation) +# - initial API and implementation +# + +source ../include/functions.sh + +CPUIDLE_STATS=./cpuidle_stats + +if [ $(id -u) != 0 ]; then + log_skip "run as non-root" + exit 0 +fi + +check_cpuidle_stats() { + trace-cmd record -e cpu_idle + trace-cmd report trace.dat > trace-cpuidle.dat + check "Running cpuidle_stats on collected data" "./$CPUIDLE_STATS" trace-cpuidle.dat +} + +check_cpuidle_stats diff --git a/cpuidle/cpuidle_04.txt b/cpuidle/cpuidle_04.txt new file mode 100644 index 0000000..8cf6bb1 --- /dev/null +++ b/cpuidle/cpuidle_04.txt @@ -0,0 +1 @@ +Run cpuidle_stats program to check runtime cpuidle statistics. diff --git a/cpuidle/cpuidle_stats.c b/cpuidle/cpuidle_stats.c new file mode 100644 index 0000000..971fcc9 --- /dev/null +++ b/cpuidle/cpuidle_stats.c @@ -0,0 +1,472 @@ +#include <getopt.h> +#include <stdio.h> +#include <stdlib.h> +#include <stdbool.h> +#include <unistd.h> +#include <string.h> +#include <values.h> +#include <sys/time.h> +#include <sys/resource.h> + +#define BUFSIZE 256 +#define MAXCSTATE 8 +#define MAX(A,B) (A > B ? A : B) +#define MIN(A,B) (A < B ? A : B) +#define AVG(A,B,I) ((A) + ((B - A) / (I))) + +static char buffer[BUFSIZE]; + +struct cpuidle_data { + double begin; + double end; + double duration; +}; + +struct cpuidle_cstate { + struct cpuidle_data *data; + int nrdata; + double avg_time; + double max_time; + double min_time; + double duration; +}; + +struct cpuidle_cstates { + struct cpuidle_cstate cstate[MAXCSTATE]; + int last_cstate; + int cstate_max; +}; + +struct cpuidle_datas { + struct cpuidle_cstates *cstates; + int nrcpus; +}; + +static inline int error(const char *str) +{ + perror(str); + return -1; +} + +static inline void *ptrerror(const char *str) +{ + perror(str); + return NULL; +} + +static int dump_data(struct cpuidle_datas *datas, int state, int count) +{ + int i = 0, j, k, nrcpus = datas->nrcpus; + struct cpuidle_cstates *cstates; + struct cpuidle_cstate *cstate; + + do { + cstates = &datas->cstates[i]; + + for (j = 0; j < cstates->cstate_max + 1; j++) { + + if (state != -1 && state != j) + continue; + + cstate = &cstates->cstate[j]; + + for (k = 0; k < MIN(count, cstate->nrdata); k++) { + printf("%lf %d\n", cstate->data[k].begin, j); + printf("%lf 0\n", cstate->data[k].end); + } + + /* add a break */ + printf("\n"); + } + + i++; + + } while (i < nrcpus && nrcpus != -1); + + return 0; +} + +static int display_data(struct cpuidle_datas *datas, int state) +{ + int i = 0, j, nrcpus = datas->nrcpus; + struct cpuidle_cstates *cstates; + struct cpuidle_cstate *cstate; + + do { + cstates = &datas->cstates[i]; + + for (j = 0; j < cstates->cstate_max + 1; j++) { + + if (state != -1 && state != j) + continue; + + cstate = &cstates->cstate[j]; + + if (nrcpus == -1) + printf("\ncluster"); + else + printf("\ncpu%d", i); + + printf("/state%d, %d hits, total %.2lfus, "\ + "avg %.2lfus, min %.2lfus, max %.2lfus", + j, cstate->nrdata, cstate->duration, + cstate->avg_time, cstate->min_time, + cstate->max_time); + } + + i++; + + } while (i < nrcpus && nrcpus != -1); + printf("\n"); + return 0; +} + +static struct cpuidle_data *intersection(struct cpuidle_data *data1, + struct cpuidle_data *data2) +{ + double begin, end; + struct cpuidle_data *data; + + begin = MAX(data1->begin, data2->begin); + end = MIN(data1->end, data2->end); + + if (begin >= end) + return NULL; + + data = malloc(sizeof(*data)); + if (!data) + return NULL; + + data->begin = begin; + data->end = end; + data->duration = end - begin; + data->duration *= 1000000; + + return data; +} + +static struct cpuidle_cstate *inter(struct cpuidle_cstate *c1, + struct cpuidle_cstate *c2) +{ + int i, j; + struct cpuidle_data *interval; + struct cpuidle_cstate *result; + struct cpuidle_data *data = NULL; + size_t index; + + if (!c1) + return c2; + if (!c2) + return c1; + + result = calloc(sizeof(*result), 1); + if (!result) + return NULL; + + for (i = 0, index = 0; i < c1->nrdata; i++) { + + for (j = index; j < c2->nrdata; j++) { + + /* intervals are ordered, no need to go further */ + if (c1->data[i].end < c2->data[j].begin) + break; + + /* primary loop begins where we ended */ + if (c1->data[i].begin > c2->data[j].end) + index = j; + + interval = intersection(&c1->data[i], &c2->data[j]); + if (!interval) + continue; + + result->min_time = MIN(!result->nrdata ? 999999.0 : + result->min_time, + interval->duration); + + result->max_time = MAX(result->max_time, + interval->duration); + + result->avg_time = AVG(result->avg_time, + interval->duration, + result->nrdata + 1); + + result->duration += interval->duration; + + result->nrdata++; + + data = realloc(data, sizeof(*data) * + (result->nrdata + 1)); + if (!data) + return NULL; + + result->data = data; + result->data[result->nrdata - 1] = *interval; + + free(interval); + } + } + + return result; +} + +static int store_data(double time, int state, int cpu, + struct cpuidle_datas *datas, int count) +{ + struct cpuidle_cstates *cstates = &datas->cstates[cpu]; + struct cpuidle_cstate *cstate; + struct cpuidle_data *data; + int nrdata, last_cstate = cstates->last_cstate; + + /* ignore when we got a "closing" state first */ + if (state == -1 && !cstates->cstate_max) + return 0; + + cstate = &cstates->cstate[state == -1 ? last_cstate : state ]; + data = cstate->data; + nrdata = cstate->nrdata; + + if (state == -1) { + + data = &data[nrdata]; + + data->end = time; + data->duration = data->end - data->begin; + + /* That happens when precision digit in the file exceed + * 7 (eg. xxx.1000000). Ignoring the result because I don't + * find a way to fix with the sscanf used in the caller + */ + if (data->duration < 0) + return 0; + + /* convert to us */ + data->duration *= 1000000; + cstate->min_time = MIN(!nrdata ? 999999.0 : cstate->min_time, + data->duration); + cstate->max_time = MAX(cstate->max_time, data->duration); + cstate->avg_time = AVG(cstate->avg_time, data->duration, + cstate->nrdata + 1); + cstate->duration += data->duration; + cstate->nrdata++; + + return 0; + } + + data = realloc(data, sizeof(*data) * (nrdata + 1)); + if (!data) + return error("realloc data");; + + data[nrdata].begin = time; + + cstates->cstate[state].data = data; + cstates->cstate_max = MAX(cstates->cstate_max, state); + cstates->last_cstate = state; + + return 0; +} + +static struct cpuidle_datas *load_data(const char *path) +{ + FILE *f; + unsigned int state = 0, cpu = 0, nrcpus= 0; + double time, begin, end; + size_t count; + + struct cpuidle_datas *datas; + + f = fopen(path, "r"); + if (!f) + return ptrerror("fopen"); + + for (count = 0; count < 2; count++) { + fgets(buffer, BUFSIZE, f); + sscanf(buffer, "cpus=%u", &nrcpus); + } + + if (!nrcpus) + return ptrerror("read error for 'cpus=' in trace file"); + + datas = malloc(sizeof(*datas)); + if (!datas) + return ptrerror("malloc datas"); + + datas->cstates = calloc(sizeof(*datas->cstates), nrcpus); + if (!datas->cstates) + return ptrerror("calloc cstate"); + + datas->nrcpus = nrcpus; + + for (; fgets(buffer, BUFSIZE, f); count++) { + + sscanf(buffer, "%*[^]]] %lf:%*[^=]=%u%*[^=]=%d", + &time, &state, &cpu); + + if (count == 2) + begin = time; + end = time; + + store_data(time, state, cpu, datas, count); + } + + fclose(f); + + fprintf(stderr, "Log is %lf secs long with %d events\n", + end - begin, count); + + return datas; +} + +struct cpuidle_datas *cluster_data(struct cpuidle_datas *datas) +{ + struct cpuidle_cstate *c1, *cstates; + struct cpuidle_datas *result; + int i, j; + int cstate_max = -1; + + result = malloc(sizeof(*result)); + if (!result) + return NULL; + + result->nrcpus = -1; /* the cluster */ + + result->cstates = calloc(sizeof(*result->cstates), 1); + if (!result->cstates) + return NULL; + + /* hack but negligeable overhead */ + for (i = 0; i < datas->nrcpus; i++) + cstate_max = MAX(cstate_max, datas->cstates[i].cstate_max); + result->cstates[0].cstate_max = cstate_max; + + for (i = 0; i < cstate_max + 1; i++) { + + for (j = 0, cstates = NULL; j < datas->nrcpus; j++) { + + c1 = &datas->cstates[j].cstate[i]; + + cstates = inter(cstates, c1); + if (!cstates) + continue; + } + + result->cstates[0].cstate[i] = *cstates; + } + + return result; +} + +static int help(const char *cmd) +{ + fprintf(stderr, "%s [-d/--dump] [-c/--cstate=x] <file>\n", cmd); + exit(0); +} + +static struct option long_options[] = { + { "dump", 0, 0, 'd' }, + { "iterations", 0, 0, 'i' }, + { "cstate", 0, 0, 'c' }, + { "debug", 0, 0, 'g' }, + { "verbose", 0, 0, 'v' }, + { "help", 0, 0, 'h' }, + { 0, 0, 0, 0 } +}; + +struct idledebug_options { + bool debug; + bool dump; + int cstate; + int iterations; +}; + +int getoptions(int argc, char *argv[], struct idledebug_options *options) +{ + int c; + + memset(options, 0, sizeof(*options)); + options->cstate = -1; + + while (1) { + + int optindex = 0; + + c = getopt_long(argc, argv, "gdvhi:c:", + long_options, &optindex); + if (c == -1) + break; + + switch (c) { + case 'g': + options->debug = true; + break; + case 'd': + options->dump = true; + break; + case 'i': + options->iterations = atoi(optarg); + break; + case 'c': + options->cstate = atoi(optarg); + break; + case 'h': + help(argv[0]); + break; + case '?': + fprintf(stderr, "%s: Unknown option %c'.\n", + argv[0], optopt); + default: + return -1; + } + } + + if (options->cstate >= MAXCSTATE) { + fprintf(stderr, "C-state must be less than %d\n", + MAXCSTATE); + return -1; + } + + if (options->iterations < 0) { + fprintf(stderr, "dump values must be a positive value\n"); + } + + if (optind == argc) { + fprintf(stderr, "expected filename\n"); + return -1; + } + + return 0; +} + +int main(int argc, char *argv[]) +{ + struct cpuidle_datas *datas; + struct cpuidle_datas *cluster; + struct idledebug_options options; + struct rusage rusage; + + + if (getoptions(argc, argv, &options)) + return 1; + + datas = load_data(argv[optind]); + if (!datas) + return 1; + + cluster = cluster_data(datas); + if (!cluster) + return 1; + + if (options.dump > 0) { + dump_data(datas, options.cstate, options.iterations); + dump_data(cluster, options.cstate, options.iterations); + } else { + display_data(datas, options.cstate); + display_data(cluster, options.cstate); + } + + if (options.debug) { + getrusage(RUSAGE_SELF, &rusage); + printf("max rss : %ld kB\n", rusage.ru_maxrss); + } + + return 0; +} -- 1.7.9.5

11 years, 7 months

← Newer
1
2
3
4
5
6
7
8
9
Older →

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

linaro-dev February 2013