lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <aJtG0xmBBgwnTANg@intel.com>
Date: Tue, 12 Aug 2025 09:51:15 -0400
From: Rodrigo Vivi <rodrigo.vivi@...el.com>
To: <zhaoguohan@...inos.cn>, Karthik Poosa <karthik.poosa@...el.com>, "Riana
 Tauro" <riana.tauro@...el.com>
CC: <lucas.demarchi@...el.com>, <thomas.hellstrom@...ux.intel.com>,
	<airlied@...il.com>, <simona@...ll.ch>, <intel-xe@...ts.freedesktop.org>,
	<dri-devel@...ts.freedesktop.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [RESEND][PATCH] drm/xe/hwmon: Return early on power limit read
 failure

On Tue, Aug 12, 2025 at 02:59:30PM +0800, zhaoguohan@...inos.cn wrote:
> From: GuoHan Zhao <zhaoguohan@...inos.cn>
> 
> In xe_hwmon_pcode_rmw_power_limit(), when xe_pcode_read() fails,
> the function logs the error but continues to execute the subsequent
> logic. This can result in undefined behavior as the values val0 and
> val1 may contain invalid data.
> 
> Fix this by adding an early return after logging the read failure,
> ensuring that we don't proceed with potentially corrupted data.
> 

Fixes: 8aa7306631f0 ("drm/xe/hwmon: Fix xe_hwmon_power_max_write")
Cc: Riana Tauro <riana.tauro@...el.com>
Cc: Karthik Poosa <karthik.poosa@...el.com>
Cc: Badal Nilawar <badal.nilawar@...el.com>

It looks like the original idea was to try to write even if the
read failed, but it was not a RMW function. Then, when it got
moved to RMW this ignored error was forgotten.

> Signed-off-by: GuoHan Zhao <zhaoguohan@...inos.cn>
> ---
>  drivers/gpu/drm/xe/xe_hwmon.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_hwmon.c b/drivers/gpu/drm/xe/xe_hwmon.c
> index f08fc4377d25..eb410c5293e7 100644
> --- a/drivers/gpu/drm/xe/xe_hwmon.c
> +++ b/drivers/gpu/drm/xe/xe_hwmon.c
> @@ -190,9 +190,11 @@ static int xe_hwmon_pcode_rmw_power_limit(const struct xe_hwmon *hwmon, u32 attr
>  						  READ_PL_FROM_PCODE : READ_PL_FROM_FW),
>  						  &val0, &val1);
>  
> -	if (ret)
> +	if (ret) {
>  		drm_dbg(&hwmon->xe->drm, "read failed ch %d val0 0x%08x, val1 0x%08x, ret %d\n",
>  			channel, val0, val1, ret);
> +			return ret;

Please change this to drm_err. Now this is an error that needs to be visible.

But also, I believe this error needs to be propagated up to the caller so anyone
using hwmon will have a clear indication that the write failed.

> +	}
>  
>  	if (attr == PL1_HWMON_ATTR)
>  		val0 = (val0 & ~clr) | set;
> -- 
> 2.43.0
> 

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ