Re: Bug#1127612: linux-image-6.12.69+deb13-amd64: hp_bioscfg mm/page_alloc.c:4802 warning

["Uwe Kleine-König" <ukleinek@debian.org>]

[-- Attachment #1: Type: text/plain, Size: 2193 bytes --]

Hello,

On Tue, Feb 17, 2026 at 07:53:17AM -0600, Mario Limonciello wrote:
> On 2/16/26 10:18 AM, Paul Kerry wrote:
> > 
> > Hi Uwe
> > 
> > Thanks for the reply and information.
> > 
> > You are correct: when I boot the HP Z2 Tower G9 into the older
> > 6.12.63+deb13-amd64 kernel, the hp_bioscfg module is not loaded which I
> > had not noticed, and running a "modprobe hp_bioscfg" on that older
> > 6.12.63+deb13-amd64 kernel does indeed generate the same dmesg warnings
> > as 6.12.69+deb13-amd64 does, which auto-loads the hp_bioscfg module.
> > 
> > I've not had any noticeable problems caused by this issue since
> > 6.12.69+deb13-amd64 came out.
> 
> Considering there appears to be a rather large number of attributes on this
> system, maybe we should be using a different memory allocation method?
> 
> Something like this:
> 
> diff --git a/drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
> b/drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
> index f346aad8e9d89..72bd6037300ab 100644
> --- a/drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
> +++ b/drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
> @@ -94,8 +94,11 @@ int hp_alloc_enumeration_data(void)
>         bioscfg_drv.enumeration_instances_count =
>                 hp_get_instance_count(HP_WMI_BIOS_ENUMERATION_GUID);
> 
> -       bioscfg_drv.enumeration_data =
> kcalloc(bioscfg_drv.enumeration_instances_count,
> - sizeof(*bioscfg_drv.enumeration_data), GFP_KERNEL);
> +       if (!bioscfg_drv.enumeration_instances_count)
> +               return -EINVAL;
> +
> +       bioscfg_drv.enumeration_data =
> kvcalloc(bioscfg_drv.enumeration_instances_count,
> + sizeof(*bioscfg_drv.enumeration_data), GFP_KERNEL);
>         if (!bioscfg_drv.enumeration_data) {
>                 bioscfg_drv.enumeration_instances_count = 0;
>                 return -ENOMEM;

The patch looks fine to me and now I wonder how to continue here. Would
a test by Paul help? Or are we sure enough the patch is fine?

@Paul: Could you test the patch? Are you able to create a kernel image
with that patch included yourself, or would a test package help you?

Best regards
Uwe

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]