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: <20241031134321.GW1279924@google.com>
Date: Thu, 31 Oct 2024 22:43:21 +0900
From: Sergey Senozhatsky <senozhatsky@...omium.org>
To: Jani Nikula <jani.nikula@...ux.intel.com>
Cc: Sergey Senozhatsky <senozhatsky@...omium.org>,
	David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>,
	Rodrigo Vivi <rodrigo.vivi@...el.com>,
	Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>,
	Tvrtko Ursulin <tursulin@...ulin.net>,
	intel-gfx@...ts.freedesktop.org, intel-xe@...ts.freedesktop.org,
	dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org
Subject: Re: [RFC][PATCH] drm: i915: do not NULL deref hdmi attached_connector

On (24/10/31 13:33), Jani Nikula wrote:
> > intel_ddi_init() may skip connector initialization, for instance,
> > both intel_ddi_init_dp_connector() and intel_ddi_init_hdmi_connector()
> > are optional.  This leads to situation that ->attached_connector may
> > be NULL for some connectors.  For instance, on my setup 'DDI A/PHY A'
> > and 'DDI TC1/PHY TC1' are not initialized.
> >
> > However, functions like intel_dp_dual_mode_set_tmds_output() and
> > friends don't take this into consideration.  This leads to NULL
> > ptr-derefs:
> >
> > KASAN: null-ptr-deref in range [0x0000000000000848-0x000000000000084f]
> > RIP: 0010:intel_hdmi_encoder_shutdown+0x105/0x230
> > Call Trace:
> > <TASK>
> > i915_driver_shutdown+0x2d8/0x490
> > pci_device_shutdown+0x83/0x150
> > device_shutdown+0x4ad/0x660
> > __se_sys_reboot+0x29c/0x4d0
> > do_syscall_64+0x60/0x90
> >
> > Add a new helper to avoid NULL ->attached_connector derefs and
> > switch some intel_hdmi function to it.  I'm not sure if we need
> > to switch all or just intel_dp_dual_mode_set_tmds_output() (I
> > have only seen this one doing NULL derefs so far).
> 
> I think the question is, what are we doing running this code if the
> connector initialization was skipped?

I'm not in position to answer that question, I guess it wasn't even
asked to me.  But...

For instance, intel_ddi_init_hdmi_connector()->intel_hdmi_init_connector()
can "error out" and leave ->attached_connector NULL; I can count 3
conditional returns before `->attached_connector = intel_connector`
assignment, yet none of the upper functions would even know, because
intel_hdmi_init_connector() returns void.  And this is not the only case.
So there are several ways to have ->attached_connector == NULL.

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ