public inbox for linux-hwmon@vger.kernel.org
 help / color / mirror / Atom feed
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
To: "Nilawar, Badal" <badal.nilawar@intel.com>
Cc: Guenter Roeck <linux@roeck-us.net>,
	Andi Shyti <andi.shyti@linux.intel.com>,
	<linux-hwmon@vger.kernel.org>, <intel-xe@lists.freedesktop.org>
Subject: Re: [Intel-xe] [PATCH v3 1/6] drm/xe/hwmon: Add HWMON infrastructure
Date: Tue, 8 Aug 2023 17:31:35 -0400	[thread overview]
Message-ID: <ZNK0NzoBlGt6chs2@intel.com> (raw)
In-Reply-To: <710b7d67-2882-a077-cbd8-ab42c41aa3ca@intel.com>

On Fri, Aug 04, 2023 at 08:06:22PM +0530, Nilawar, Badal wrote:
> 
> 
> On 04-08-2023 19:56, Guenter Roeck wrote:
> > On 8/4/23 06:19, Nilawar, Badal wrote:
> > > 
> > > Hi Guenter,
> > > On 03-08-2023 04:42, Guenter Roeck wrote:
> > > > On 8/2/23 15:40, Andi Shyti wrote:
> > > > > Hi Badal,
> > > > > 
> > > > > [...]
> > > > > 
> > > > > > +struct xe_hwmon_data {
> > > > > > +    struct device *hwmon_dev;
> > > > > > +    struct xe_gt *gt;
> > > > > > +    char name[12];
> > > > > > +};
> > > > > > +
> > > > > > +struct xe_hwmon {
> > > > > > +    struct xe_hwmon_data ddat;
> > > > > > +    struct mutex hwmon_lock;
> > > > > > +};
> > > > > 
> > > > > why do we need two structures here? Can we merge them?
> > > > > 
> > > > 
> > > > A later patch adds multiple hwmon devices which makes use of it.
> > > > I think that is flawed, and I am not inclined to accept it.
> > > Is there any obvious reason that there shouldn't be multiple
> > > devices? In i915 we are doing the same.
> > > https://patchwork.freedesktop.org/patch/497324/?series=104278&rev=3
> > > 
> > 
> > Technically you can do whatever you like as long as the code doesn't reside
> > in drivers/hwmon. I won't NACK it, but I won't give it a Reviewed-by:
> > either. i915 shouldn't do it, but I didn't realize what they are doing
> > at the time. Other drivers doing it wrong is not an argument. You can't
> > argue that you may drive faster than the speed limit because others do it
> > or because police didn't stop you last time you did either.
> > 
> > One chip, one hwmon device. Do you have separate parent devices for
> > all your hwmon devices ? If yes, you can argue that having multiple hwmon
> > devices make sense. If not, you can't.
> Thanks for clarification. There is only one parent device. So will try to
> accommodate single hwmon device.

Well, it is one PCI device, but under 1 pci device we can have multiple "tiles"
that can duplicate many components. Inside each tile we can even have multiple
"gt"s.

But back to the tile, each tile has its own metrics. It's own power delivery,
own sensors and all. They can even be seen as independent devices from this
angle.

I'm afraid that the attempt to put everything as one device, but all the
entries duplicated per tile/gt we might end up with a messed api.

> 
> Regards,
> Badal
> > 
> > Guenter
> > 

  reply	other threads:[~2023-08-08 21:31 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-02 13:52 [PATCH v3 0/6] Add HWMON support for DGFX Badal Nilawar
2023-08-02 13:52 ` [PATCH v3 1/6] drm/xe/hwmon: Add HWMON infrastructure Badal Nilawar
2023-08-02 14:15   ` Guenter Roeck
2023-08-02 22:40   ` Andi Shyti
2023-08-02 23:11     ` Guenter Roeck
2023-08-02 23:34       ` Andi Shyti
2023-08-03  0:06         ` Guenter Roeck
2023-08-02 23:12     ` Guenter Roeck
2023-08-04 13:19       ` Nilawar, Badal
2023-08-04 14:26         ` Guenter Roeck
2023-08-04 14:36           ` Nilawar, Badal
2023-08-08 21:31             ` Rodrigo Vivi [this message]
2023-08-08 22:07               ` [Intel-xe] " Guenter Roeck
2023-08-11 16:01                 ` Rodrigo Vivi
2023-08-11 17:39                   ` Guenter Roeck
2023-08-11 18:48                     ` Rodrigo Vivi
2023-08-04 14:43           ` Nilawar, Badal
2023-08-04 13:25     ` Nilawar, Badal
2023-08-02 13:52 ` [PATCH v3 2/6] drm/xe/hwmon: Expose power attributes Badal Nilawar
2023-08-02 23:23   ` Andi Shyti
2023-08-04 14:21     ` Nilawar, Badal
2023-08-02 13:52 ` [PATCH v3 3/6] drm/xe/hwmon: Expose card reactive critical power Badal Nilawar
2023-08-02 23:28   ` Andi Shyti
2023-08-04 13:31     ` Nilawar, Badal
2023-08-02 13:52 ` [PATCH v3 4/6] drm/xe/hwmon: Expose input voltage attribute Badal Nilawar
2023-08-02 23:32   ` Andi Shyti
2023-08-04 13:30     ` Nilawar, Badal
2023-08-02 13:52 ` [PATCH v3 5/6] drm/xe/hwmon: Expose hwmon energy attribute Badal Nilawar
2023-08-02 14:14   ` Guenter Roeck
2023-08-03  6:34     ` Nilawar, Badal
2023-08-03 14:42       ` Guenter Roeck
2023-08-02 13:52 ` [PATCH v3 6/6] drm/xe/hwmon: Expose power1_max_interval Badal Nilawar

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=ZNK0NzoBlGt6chs2@intel.com \
    --to=rodrigo.vivi@intel.com \
    --cc=andi.shyti@linux.intel.com \
    --cc=badal.nilawar@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=linux-hwmon@vger.kernel.org \
    --cc=linux@roeck-us.net \
    /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