From: Thierry Reding <thierry.reding-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
To: Wei Ni <wni-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
Cc: rui.zhang-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org,
mikko.perttunen-/1wQRMveznE@public.gmane.org,
swarren-3lzwWm7+Weoh9ZMKESR00Q@public.gmane.org,
linux-tegra-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: [PATCH V1 05/10] thermal: tegra: add sysfs to dump registers
Date: Wed, 13 Jan 2016 16:17:40 +0100 [thread overview]
Message-ID: <20160113151740.GE2588@ulmo> (raw)
In-Reply-To: <1452672007-330-1-git-send-email-wni-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
[-- Attachment #1: Type: text/plain, Size: 5637 bytes --]
On Wed, Jan 13, 2016 at 04:00:07PM +0800, Wei Ni wrote:
> Add sysfs interface to dump registers for debug.
You're adding a debugfs interface here, not sysfs.
> diff --git a/drivers/thermal/tegra/tegra_soctherm.c b/drivers/thermal/tegra/tegra_soctherm.c
[...]
> +#ifdef CONFIG_DEBUG_FS
> +static int regs_show(struct seq_file *s, void *data)
> +{
> + struct platform_device *pdev = s->private;
> + struct tegra_soctherm *ts = platform_get_drvdata(pdev);
> + struct tegra_tsensor *tsensors = ts->tsensors;
> + u32 r, state;
> + int i;
> +
> + seq_puts(s, "-----TSENSE (convert HW)-----\n");
> +
> + for (i = 0; tsensors[i].name; i++) {
> + r = readl(ts->regs + tsensors[i].base + SENSOR_CONFIG1);
> + state = REG_GET_MASK(r, SENSOR_CONFIG1_TEMP_ENABLE);
> + if (!state)
> + continue;
> +
> + seq_printf(s, "%s: ", tsensors[i].name);
> +
> + seq_printf(s, "En(%d) ", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG1_TIDDQ_EN_MASK);
> + seq_printf(s, "tiddq(%d) ", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG1_TEN_COUNT_MASK);
> + seq_printf(s, "ten_count(%d) ", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG1_TSAMPLE_MASK);
> + seq_printf(s, "tsample(%d) ", state + 1);
> +
> + r = readl(ts->regs + tsensors[i].base + SENSOR_STATUS1);
> + state = REG_GET_MASK(r, SENSOR_STATUS1_TEMP_VALID_MASK);
> + seq_printf(s, "Temp(%d/", state);
> + state = REG_GET_MASK(r, SENSOR_STATUS1_TEMP_MASK);
> + seq_printf(s, "%d) ", translate_temp(state));
> +
> + r = readl(ts->regs + tsensors[i].base + SENSOR_STATUS0);
> + state = REG_GET_MASK(r, SENSOR_STATUS0_VALID_MASK);
> + seq_printf(s, "Capture(%d/", state);
> + state = REG_GET_MASK(r, SENSOR_STATUS0_CAPTURE_MASK);
> + seq_printf(s, "%d) ", state);
> +
> + r = readl(ts->regs + tsensors[i].base + SENSOR_CONFIG0);
> + state = REG_GET_MASK(r, SENSOR_CONFIG0_STOP);
> + seq_printf(s, "Stop(%d) ", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG0_TALL_MASK);
> + seq_printf(s, "Tall(%d) ", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG0_TCALC_OVER);
> + seq_printf(s, "Over(%d/", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG0_OVER);
> + seq_printf(s, "%d/", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG0_CPTR_OVER);
> + seq_printf(s, "%d) ", state);
> +
> + r = readl(ts->regs + tsensors[i].base + SENSOR_CONFIG2);
> + state = REG_GET_MASK(r, SENSOR_CONFIG2_THERMA_MASK);
> + seq_printf(s, "Therm_A/B(%d/", state);
> + state = REG_GET_MASK(r, SENSOR_CONFIG2_THERMB_MASK);
> + seq_printf(s, "%d)\n", (s16)state);
> + }
This isn't really a register dump, it's a decoded form of the register
contents. I think it'd be better to either provide a debugfs file with
the raw register contents, perhaps with register name and offset, or a
decoded version. Or both.
> +
> + r = readl(ts->regs + SENSOR_PDIV);
> + seq_printf(s, "PDIV: 0x%x\n", r);
> +
> + r = readl(ts->regs + SENSOR_HOTSPOT_OFF);
> + seq_printf(s, "HOTSPOT: 0x%x\n", r);
> +
> + seq_puts(s, "\n");
> + seq_puts(s, "-----SOC_THERM-----\n");
I don't think these separators are any good. debugfs files are
essentially for free, so if you have logical separation of the content,
as indicated by the separator here, a better option would be to make
these separate files.
> + r = readl(ts->regs + SENSOR_TEMP1);
> + state = REG_GET_MASK(r, SENSOR_TEMP1_CPU_TEMP_MASK);
> + seq_printf(s, "Temperatures: CPU(%d) ", translate_temp(state));
> + state = REG_GET_MASK(r, SENSOR_TEMP1_GPU_TEMP_MASK);
> + seq_printf(s, " GPU(%d) ", translate_temp(state));
> + r = readl(ts->regs + SENSOR_TEMP2);
> + state = REG_GET_MASK(r, SENSOR_TEMP2_PLLX_TEMP_MASK);
> + seq_printf(s, " PLLX(%d) ", translate_temp(state));
> + state = REG_GET_MASK(r, SENSOR_TEMP2_MEM_TEMP_MASK);
> + seq_printf(s, " MEM(%d)\n", translate_temp(state));
Isn't this something that the thermal core support already exports via
some interface?
> +static int soctherm_debug_init(struct platform_device *pdev)
> +{
> + struct dentry *tegra_soctherm_root;
> +
> + tegra_soctherm_root = debugfs_create_dir("tegra_soctherm", NULL);
You should check for potential failure here, otherwise you might end up
in a situation where you create the soctherm files in the debugfs root,
which is almost certainly not what you want.
> + debugfs_create_file("regs", 0644, tegra_soctherm_root,
> + pdev, ®s_fops);
> +
> + return 0;
> +}
> +#else
> +static inline int soctherm_debug_init(struct platform_device *pdev)
> +{ return 0; }
> +#endif
> +
> int tegra_soctherm_probe(struct platform_device *pdev,
> struct tegra_tsensor *tsensors,
> const struct tegra_tsensor_group **ttgs,
> @@ -154,6 +276,10 @@ int tegra_soctherm_probe(struct platform_device *pdev,
> if (!tegra)
> return -ENOMEM;
>
> + dev_set_drvdata(&pdev->dev, tegra);
> +
> + tegra->tsensors = tsensors;
> +
> res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> tegra->regs = devm_ioremap_resource(&pdev->dev, res);
> if (IS_ERR(tegra->regs))
> @@ -243,6 +369,8 @@ int tegra_soctherm_probe(struct platform_device *pdev,
> tegra->thermctl_tzs[ttgs[i]->id] = tz;
> }
>
> + soctherm_debug_init(pdev);
If you never check and don't care about the return value, I suggest
making the soctherm_debugfs_init() function return void. You should
still check for errors within the function to make sure you're not
doing anything you shouldn't, but you can silently ignore errors if
you don't care, or perhaps give an indication as to what failed with
a dev_warn() or similar.
Thierry
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
next prev parent reply other threads:[~2016-01-13 15:17 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-01-13 8:00 [PATCH V1 05/10] thermal: tegra: add sysfs to dump registers Wei Ni
[not found] ` <1452672007-330-1-git-send-email-wni-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
2016-01-13 15:17 ` Thierry Reding [this message]
2016-01-14 11:01 ` Wei Ni
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=20160113151740.GE2588@ulmo \
--to=thierry.reding-re5jqeeqqe8avxtiumwx3w@public.gmane.org \
--cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=linux-tegra-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=mikko.perttunen-/1wQRMveznE@public.gmane.org \
--cc=rui.zhang-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org \
--cc=swarren-3lzwWm7+Weoh9ZMKESR00Q@public.gmane.org \
--cc=wni-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org \
/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