Linux Hardware Monitor development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Akshay Gupta" <Akshay.Gupta@amd.com>,
	linux-hwmon@vger.kernel.org, devicetree@vger.kernel.org
Cc: robh@kernel.org, conor+dt@kernel.org
Subject: Re: [PATCH v1 1/4] misc: amd-sbi: use guard() for data->lock mutex
Date: Mon, 16 Nov 2026 12:15:25 +0000	[thread overview]
Message-ID: <sashiko-outbox-217000@kernel.org> (raw)
In-Reply-To: <20261116120228.865196-2-Akshay.Gupta@amd.com>

Thank you for your contribution! Sashiko AI review found no new issues
in this patch, but noticed 1 related pre-existing issue it does not
introduce:
- [High] misc: amd-sbi: uninitialized ec and error overwrite in
  rmi_mailbox_xfer() [1]
--

commit 8e0c735b23732b6b78100642d3a437c0cd3041c0
Author: Prathima Lk <Prathima.Lk@amd.com>

misc: amd-sbi: use guard() for data->lock mutex

This commit converts manual mutex_lock() and mutex_unlock() calls on
data->lock in drivers/misc/amd-sbi/rmi-core.c to guard() and
scoped_guard(), simplifying the error-return paths.

> diff --git a/drivers/misc/amd-sbi/rmi-core.c b/drivers/misc/amd-sbi/rmi-core.c
> index 95c9109101b53..277f121409728 100644
> --- a/drivers/misc/amd-sbi/rmi-core.c
> +++ b/drivers/misc/amd-sbi/rmi-core.c

[ ... ]

> @@ -425,29 +425,27 @@ int rmi_mailbox_xfer(struct sbrmi_data *data,
>  	ret = regmap_read_poll_timeout(data->regmap, SBRMI_STATUS, sw_status,
>  				       sw_status & SW_ALERT_MASK, 500, 2000000);
>  	if (ret)
> -		goto exit_unlock;
> +		return ret;
>  
>  	ret = regmap_read(data->regmap, SBRMI_OUTBNDMSG7, &ec);
> -	if (ret || ec)
> -		goto exit_clear_alert;
> -
> -	/* Clear the input value before updating the output data */
> -	msg->mb_in_out = 0;
> -
> -	/*
> -	 * For a read operation, the initiator (BMC) reads the firmware
> -	 * response Command Data Out[31:0] from SBRMI::OutBndMsg_inst[4:1]
> -	 * {SBRMI_x34(MSB):SBRMI_x31(LSB)}.
> -	 */
> -	for (i = 0; i < AMD_SBI_MB_DATA_SIZE; i++) {
> -		ret = regmap_read(data->regmap,
> -				  SBRMI_OUTBNDMSG1 + i, &bytes);
> -		if (ret < 0)
> -			break;
> -		msg->mb_in_out |= bytes << i * 8;
> +	if (!ret && !ec) {
> +		/* Clear the input value before updating the output data */
> +		msg->mb_in_out = 0;
> +
> +		/*
> +		 * For a read operation, the initiator (BMC) reads the firmware
> +		 * response Command Data Out[31:0] from SBRMI::OutBndMsg_inst[4:1]
> +		 * {SBRMI_x34(MSB):SBRMI_x31(LSB)}.
> +		 */
> +		for (i = 0; i < AMD_SBI_MB_DATA_SIZE; i++) {
> +			ret = regmap_read(data->regmap,
> +					  SBRMI_OUTBNDMSG1 + i, &bytes);
> +			if (ret < 0)
> +				break;
> +			msg->mb_in_out |= bytes << i * 8;
> +		}
>  	}
>  
> -exit_clear_alert:
>  	/*
>  	 * BMC must write 1'b1 to SBRMI::Status[SwAlertSts] to clear the
>  	 * ALERT to initiator
> @@ -458,8 +456,7 @@ int rmi_mailbox_xfer(struct sbrmi_data *data,
>  		ret = -EPROTOTYPE;
>  		msg->fw_ret_code = ec;
>  	}
> -exit_unlock:
> -	mutex_unlock(&data->lock);
> +
>  	return ret;
>  }

[Severity: High]
This problem wasn't introduced by this patch [1], but what happens
if regmap_read() fails when reading SBRMI_OUTBNDMSG7 or
SBRMI_OUTBNDMSG1..4?

ec is declared on the stack without initialization in rmi_mailbox_xfer().
If reading SBRMI_OUTBNDMSG7 fails, ec remains uninitialized when checked
at "if (ec)" at the end of the function; if it happens to be non-zero, the
uninitialized value is stored into msg->fw_ret_code and copied to
userspace by apml_mailbox_xfer().

In addition, the regmap_write() to SBRMI_STATUS unconditionally overwrites
ret, so a failed read of SBRMI_OUTBNDMSG7 (when ec happens to be 0) or
SBRMI_OUTBNDMSG1..4 returns 0 when clearing the alert succeeds, leaving
callers with stale or partial msg->mb_in_out data.

[1] https://sashiko.dev/#/bug/linux-37f3c6ab-ea0e-4ada-a81e-e6c232558ae3

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261116120228.865196-1-Akshay.Gupta@amd.com?part=1

  reply	other threads:[~2026-11-16 12:15 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-16 12:02 [PATCH v1 0/4] misc: amd-sbi: add DIMM thermal sensors to sbrmi Akshay Gupta
2026-11-16 12:02 ` [PATCH v1 1/4] misc: amd-sbi: use guard() for data->lock mutex Akshay Gupta
2026-11-16 12:15   ` sashiko-bot [this message]
2026-11-16 12:02 ` [PATCH v1 2/4] dt-bindings: hwmon: add DIMM thermal sensor properties Akshay Gupta
2026-11-16 12:11   ` sashiko-bot
2026-11-16 12:02 ` [PATCH v1 3/4] misc: amd-sbi: add DIMM thermal sensor support Akshay Gupta
2026-11-16 12:13   ` sashiko-bot
2026-11-16 12:02 ` [PATCH v1 4/4] Documentation: hwmon: document DIMM thermal sensor channels Akshay Gupta
2026-11-16 12:08   ` 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-217000@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=Akshay.Gupta@amd.com \
    --cc=conor+dt@kernel.org \
    --cc=devicetree@vger.kernel.org \
    --cc=linux-hwmon@vger.kernel.org \
    --cc=robh@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