linux-kernel - Re: [syzbot] riscv/fixes test error: can't ssh into the instance (2)

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: <fcbb98be-3a15-4a24-96ab-74d169337983@app.fastmail.com>
Date: Wed, 21 May 2025 21:51:55 +0200
From: "Arnd Bergmann" <arnd@...db.de>
To: "Alexandre Ghiti" <alex@...ti.fr>,
 syzbot <syzbot+2cae92ded758083f5bde@...kaller.appspotmail.com>,
 "Albert Ou" <aou@...s.berkeley.edu>, linux-kernel@...r.kernel.org,
 linux-riscv@...ts.infradead.org, "Palmer Dabbelt" <palmer@...belt.com>,
 "Paul Walmsley" <paul.walmsley@...ive.com>, syzkaller-bugs@...glegroups.com,
 "Mark Pearson" <mpearson-lenovo@...ebb.ca>,
 Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>,
 linux-acpi@...r.kernel.org
Subject: Re: [syzbot] riscv/fixes test error: can't ssh into the instance (2)

On Mon, May 19, 2025, at 18:35, Alexandre Ghiti wrote:
> On 5/19/25 12:12, syzbot wrote:
>
> Because we don't have acpi enabled and then acpi_kobj is null 
> (https://elixir.bootlin.com/linux/v6.15-rc6/source/fs/sysfs/group.c#L131).
>
> The following patch fixes it, but not sure this is the right way, let me 
> know if I should send it, it would be nice to have it in 6.15:
>

> --- a/drivers/acpi/platform_profile.c
> +++ b/drivers/acpi/platform_profile.c
> @@ -688,6 +688,9 @@ static int __init platform_profile_init(void)
>   {
>          int err;
>
> +       if (acpi_disabled)
> +               return -ENOTSUPP;
> +
>          err = class_register(&platform_profile_class);
>          if (err)

The change makes sense to me.

I double-checked that returning an error from the initcall
causes no problems, but it is only used for printing it with trace_initcall_finish_cb().

As far as I can tell, the problem has existed in theory since
linux-6.14, when the sysfs_create_group() call was moved
from platform_profile_register() to the module_init call.

Maybe clarify this and add

Fixes: 77be5cacb2c2 ("ACPI: platform_profile: Create class for ACPI platform profile")

      Arnd