From: Chanwoo Choi <cw00.choi@samsung.com>
To: Dmitry Osipenko <digetx@gmail.com>,
linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org
Cc: leonard.crestez@nxp.com, lukasz.luba@arm.com,
a.swigon@samsung.com, m.szyprowski@samsung.com,
enric.balletbo@collabora.com, hl@rock-chips.com,
bjorn.andersson@linaro.org, jcrouse@codeaurora.org,
chanwoo@kernel.org, myungjoo.ham@samsung.com,
kyungmin.park@samsung.com
Subject: Re: [PATCH v3] PM / devfreq: Add debugfs support with devfreq_summary file
Date: Tue, 14 Jan 2020 11:23:05 +0900 [thread overview]
Message-ID: <9635588e-66c2-14d5-9d84-43dbce976722@samsung.com> (raw)
In-Reply-To: <20ebd8e0-f74c-335c-4af8-35c3dc81902f@gmail.com>
On 1/14/20 10:45 AM, Dmitry Osipenko wrote:
> 14.01.2020 04:45, Chanwoo Choi пишет:
>> Add debugfs interface to provide debugging information of devfreq device.
>> It contains 'devfreq_summary' entry to show the summary of registered
>> devfreq devices as following and the additional debugfs file will be added.
>> - /sys/kernel/debug/devfreq/devfreq_summary
>>
>> [Detailed description of each field of 'devfreq_summary' debugfs file]
>> - dev_name : Device name of h/w.
>> - dev : Device name made by devfreq core.
>> - parent_dev : If devfreq device uses the passive governor,
>> show parent devfreq device name. Otherwise, show 'null'.
>> - governor : Devfreq governor.
>> - polling_ms : If devfreq device uses the simple_ondemand governor,
>> polling_ms is necessary for the period. (unit: millisecond)
>> - cur_freq_Hz : Current Frequency (unit: Hz)
>> - old_freq_Hz : Frequency before changing. (unit: Hz)
>> - new_freq_Hz : Frequency after changed. (unit: Hz)
>>
>> [For example on Exynos5422-based Odroid-XU3 board]
>> $ cat /sys/kernel/debug/devfreq/devfreq_summary
>> dev_name dev parent_dev governor polling_ms cur_freq_Hz min_freq_Hz max_freq_Hz
>> ------------------------------ ---------- ---------- --------------- ---------- ------------ ------------ ------------
>> 10c20000.memory-controller devfreq0 null simple_ondemand 0 165000000 165000000 825000000
>> soc:bus_wcore devfreq1 null simple_ondemand 50 532000000 88700000 532000000
>> soc:bus_noc devfreq2 devfreq1 passive 0 111000000 66600000 111000000
>> soc:bus_fsys_apb devfreq3 devfreq1 passive 0 222000000 111000000 222000000
>> soc:bus_fsys devfreq4 devfreq1 passive 0 200000000 75000000 200000000
>> soc:bus_fsys2 devfreq5 devfreq1 passive 0 200000000 75000000 200000000
>> soc:bus_mfc devfreq6 devfreq1 passive 0 333000000 83250000 333000000
>> soc:bus_gen devfreq7 devfreq1 passive 0 266000000 88700000 266000000
>> soc:bus_peri devfreq8 devfreq1 passive 0 66600000 66600000 66600000
>> soc:bus_g2d devfreq9 devfreq1 passive 0 333000000 83250000 333000000
>> soc:bus_g2d_acp devfreq10 devfreq1 passive 0 266000000 66500000 266000000
>> soc:bus_jpeg devfreq11 devfreq1 passive 0 300000000 75000000 300000000
>> soc:bus_jpeg_apb devfreq12 devfreq1 passive 0 166500000 83250000 166500000
>> soc:bus_disp1_fimd devfreq13 devfreq1 passive 0 200000000 120000000 200000000
>> soc:bus_disp1 devfreq14 devfreq1 passive 0 300000000 120000000 300000000
>> soc:bus_gscl_scaler devfreq15 devfreq1 passive 0 300000000 150000000 300000000
>> soc:bus_mscl devfreq16 devfreq1 passive 0 666000000 84000000 666000000
>>
>> [lkp: Reported the build error]
>> Reported-by: kbuild test robot <lkp@intel.com>
>> Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
>> ---
>> Changes from v2:
>> - Show 'null' at 'parent_dev' field when governor of devfreq device
>> is not passive
>> Changes from v1:
>> - Drop the patch about 'devfreq_transitions' debugfs file
>> - Modify from 'hz' to 'Hz'
>> - Edit the indentation of 'devfreq_summary' when show summary
>> - Exchange sequence between PTR_ERR and IS_ERR when debugfs_create_dir
>>
>> drivers/devfreq/devfreq.c | 84 +++++++++++++++++++++++++++++++++++++++
>> 1 file changed, 84 insertions(+)
>>
>> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
>> index 89260b17598f..c5ef2d194b1b 100644
>> --- a/drivers/devfreq/devfreq.c
>> +++ b/drivers/devfreq/devfreq.c
>> @@ -10,6 +10,7 @@
>> #include <linux/kernel.h>
>> #include <linux/kmod.h>
>> #include <linux/sched.h>
>> +#include <linux/debugfs.h>
>> #include <linux/errno.h>
>> #include <linux/err.h>
>> #include <linux/init.h>
>> @@ -33,6 +34,7 @@
>> #define HZ_PER_KHZ 1000
>>
>> static struct class *devfreq_class;
>> +static struct dentry *devfreq_debugfs;
>>
>> /*
>> * devfreq core provides delayed work based load monitoring helper
>> @@ -1643,6 +1645,79 @@ static struct attribute *devfreq_attrs[] = {
>> };
>> ATTRIBUTE_GROUPS(devfreq);
>>
>> +/**
>> + * devfreq_summary_show() - Show the summary of the devfreq devices
>> + * @s: seq_file instance to show the summary of devfreq devices
>> + * @data: not used
>> + *
>> + * Show the summary of the devfreq devices via 'devfreq_summary' debugfs file.
>> + * It helps that user can know the detailed information of the devfreq devices.
>> + *
>> + * Return 0 always because it shows the information without any data change.
>> + */
>> +static int devfreq_summary_show(struct seq_file *s, void *data)
>> +{
>> + struct devfreq *devfreq;
>> + struct devfreq *p_devfreq = NULL;
>> + unsigned long cur_freq, min_freq, max_freq;
>> + unsigned int polling_ms;
>> +
>> + seq_printf(s, "%-30s %-10s %-10s %-15s %10s %12s %12s %12s\n",
>> + "dev_name",
>> + "dev",
>> + "parent_dev",
>> + "governor",
>> + "polling_ms",
>> + "cur_freq_Hz",
>> + "min_freq_Hz",
>> + "max_freq_Hz");
>> + seq_printf(s, "%30s %10s %10s %15s %10s %12s %12s %12s\n",
>> + "------------------------------",
>> + "----------",
>> + "----------",
>> + "---------------",
>> + "----------",
>> + "------------",
>> + "------------",
>> + "------------");
>> +
>> + mutex_lock(&devfreq_list_lock);
>> +
>> + list_for_each_entry_reverse(devfreq, &devfreq_list, node) {
>> + if (!strncmp(devfreq->governor_name, DEVFREQ_GOV_PASSIVE,
>> + DEVFREQ_NAME_LEN)) {
>> + struct devfreq_passive_data *data = devfreq->data;
>> +
>> + if (data)
>> + p_devfreq = data->parent;
>> + } else {
>> + p_devfreq = NULL;
>> + }
>> +
>> + mutex_lock(&devfreq->lock);
>> + cur_freq = devfreq->previous_freq,
>> + get_freq_range(devfreq, &min_freq, &max_freq);
>> + polling_ms = devfreq->profile->polling_ms,
>> + mutex_unlock(&devfreq->lock);
>> +
>> + seq_printf(s,
>> + "%-30s %-10s %-10s %-15s %10d %12ld %12ld %12ld\n",
>> + dev_name(devfreq->dev.parent),
>> + dev_name(&devfreq->dev),
>> + p_devfreq ? dev_name(&p_devfreq->dev) : "null",
>> + devfreq->governor_name,
>> + polling_ms,
>> + cur_freq,
>> + min_freq,
>> + max_freq);
>> + }
>> +
>> + mutex_unlock(&devfreq_list_lock);
>> +
>> + return 0;
>> +}
>> +DEFINE_SHOW_ATTRIBUTE(devfreq_summary);
>> +
>> static int __init devfreq_init(void)
>> {
>> devfreq_class = class_create(THIS_MODULE, "devfreq");
>> @@ -1659,6 +1734,15 @@ static int __init devfreq_init(void)
>> }
>> devfreq_class->dev_groups = devfreq_groups;
>>
>> +
> If you'll you'll take a look at [2], you may notice notice that it
> checks whether devfreq_debugfs IS_ERR [3] and then bails out.
>
> [2] https://protect2.fireeye.com/url?k=a5047d02-f8ca7cd1-a505f64d-000babff317b-ed64c541cd573190&u=https://elixir.bootlin.com/linux/v5.5-rc6/source/fs/debugfs/inode.c#L432
> [3] https://protect2.fireeye.com/url?k=0a43fffa-578dfe29-0a4274b5-000babff317b-1c1c1062bfaeb0fc&u=https://elixir.bootlin.com/linux/v5.5-rc6/source/fs/debugfs/inode.c#L316
>
> Thus you could simply remove the above error handling, making code to
> look cleaner.
>
>
OK. I'll modify it as following: Do you agree?
devfreq_debugfs = debugfs_create_dir("devfreq", NULL);
debugfs_create_file("devfreq_summary", 0444,
devfreq_debugfs, NULL,
&devfreq_summary_fops);
--
Best Regards,
Chanwoo Choi
Samsung Electronics
next prev parent reply other threads:[~2020-01-14 2:16 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CGME20200114013826epcas1p23a4c6c39f043dd09d33d48541288c1bd@epcas1p2.samsung.com>
2020-01-14 1:45 ` [PATCH v3] PM / devfreq: Add debugfs support with devfreq_summary file Chanwoo Choi
2020-01-14 1:45 ` Dmitry Osipenko
2020-01-14 2:23 ` Chanwoo Choi [this message]
2020-01-14 2:51 ` Dmitry Osipenko
2020-01-14 4:02 ` Chanwoo Choi
2020-01-14 4:05 ` Chanwoo Choi
[not found] <CGME20200107015549epcas1p28a8f26855da65790c13d843506a0b15e@epcas1p2.samsung.com>
2020-01-07 2:02 ` Chanwoo Choi
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=9635588e-66c2-14d5-9d84-43dbce976722@samsung.com \
--to=cw00.choi@samsung.com \
--cc=a.swigon@samsung.com \
--cc=bjorn.andersson@linaro.org \
--cc=chanwoo@kernel.org \
--cc=digetx@gmail.com \
--cc=enric.balletbo@collabora.com \
--cc=hl@rock-chips.com \
--cc=jcrouse@codeaurora.org \
--cc=kyungmin.park@samsung.com \
--cc=leonard.crestez@nxp.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=lukasz.luba@arm.com \
--cc=m.szyprowski@samsung.com \
--cc=myungjoo.ham@samsung.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;
as well as URLs for NNTP newsgroup(s).