Linux Hardware Monitor development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Felix Radensky" <fradensky@nvidia.com>
Cc: linux-hwmon@vger.kernel.org
Subject: Re: [PATCH 1/3] hwmon: (pmbus/core) Invalidate cached page and phase on failed change
Date: Sun, 22 Nov 2026 07:48:16 +0000	[thread overview]
Message-ID: <sashiko-outbox-227794@kernel.org> (raw)
In-Reply-To: <20261122073322.1712843-2-fradensky@nvidia.com>

Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [Medium] hwmon: (pmbus/core) pmbus_set_page() updates cache outside
  conditional blocks
--

commit 03751a48b86432f4503eb9a4970150b29ef4ab2c
Author: Felix Radensky <fradensky@nvidia.com>

hwmon: (pmbus/core) Invalidate cached page and phase on failed change

This commit resets data->currpage and data->currphase to -1 in
pmbus_set_page() before issuing PMBUS_PAGE and PMBUS_PHASE commands so
that a failed transfer forces a retry on the next access.

> diff --git a/drivers/hwmon/pmbus/pmbus_core.c b/drivers/hwmon/pmbus/pmbus_core.c
> index 82dac2c6c579f..dae56e2e6f21f 100644
> --- a/drivers/hwmon/pmbus/pmbus_core.c
> +++ b/drivers/hwmon/pmbus/pmbus_core.c
> @@ -227,6 +227,9 @@ int pmbus_set_page(struct i2c_client *client, int page, int phase)
>  
>  	if (!(data->info->func[page] & PMBUS_PAGE_VIRTUAL) &&
>  	    data->info->pages > 1 && page != data->currpage) {
> +		/* A failed page change can leave the chip on any page */
> +		data->currpage = -1;
> +
>  		pmbus_wait(client);
>  		rv = i2c_smbus_write_byte_data(client, PMBUS_PAGE, page);
>  		pmbus_update_ts(client, PMBUS_OP_WRITE | PMBUS_OP_PAGE_CHANGE);
> @@ -246,6 +249,9 @@ int pmbus_set_page(struct i2c_client *client, int page, int phase)
>  
>  	if (data->info->phases[page] && data->currphase != phase &&
>  	    !(data->info->func[page] & PMBUS_PHASE_VIRTUAL)) {
> +		/* A failed phase change can leave the chip on any phase */
> +		data->currphase = -1;
> +
>  		pmbus_wait(client);
>  		rv = i2c_smbus_write_byte_data(client, PMBUS_PHASE,
>  					       phase);

[Severity: Medium]
Should data->currpage = page and data->currphase = phase be moved inside
their respective conditional blocks?

Because both assignments remain outside the if blocks:

drivers/hwmon/pmbus/pmbus_core.c:pmbus_set_page() {
	...
	data->currpage = page;

	if (data->info->phases[page] && data->currphase != phase &&
	    !(data->info->func[page] & PMBUS_PHASE_VIRTUAL)) {
		...
	}
	data->currphase = phase;
}

any call that skips the SMBus write still overwrites the cached value
(including an invalidated -1) without programming the chip.  For example,
on a multi-page device where page 0 has phases and page 1 has
info->phases[1] == 0 (such as tps53681), if a phase change to 0xff on page
0 fails (leaving currphase == -1), an access to page 1 with phase 0xff
skips the PMBUS_PHASE write but sets currphase = 0xff.  A subsequent access
to page 0 with phase 0xff then sees currphase == phase and skips sending
PMBUS_PHASE, accessing the wrong phase on the chip.  The same happens for
currpage when accessing a page with PMBUS_PAGE_VIRTUAL set.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261122073322.1712843-1-fradensky@nvidia.com?part=1

  reply	other threads:[~2026-11-22  7:48 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-22  7:33 [PATCH 0/3] hwmon: (pmbus) Add PMBUS_FORCE_PAGE_CHANGE flag Felix Radensky
2026-11-22  7:33 ` [PATCH 1/3] hwmon: (pmbus/core) Invalidate cached page and phase on failed change Felix Radensky
2026-11-22  7:48   ` sashiko-bot [this message]
2026-11-22  7:33 ` [PATCH 2/3] hwmon: (pmbus/core) Add PMBUS_FORCE_PAGE_CHANGE flag Felix Radensky
2026-11-22  7:33 ` [PATCH 3/3] hwmon: (pmbus/mp2975) Use PMBUS_FORCE_PAGE_CHANGE Felix Radensky
2026-11-22  7:48 ` [PATCH 0/3] hwmon: (pmbus) Add PMBUS_FORCE_PAGE_CHANGE flag sashiko-bot

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=sashiko-outbox-227794@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=fradensky@nvidia.com \
    --cc=linux-hwmon@vger.kernel.org \
    --cc=sashiko-reviews@lists.linux.dev \
    /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