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: <Y7hogpIPCmvrX+1a@phenom.ffwll.local>
Date:   Fri, 6 Jan 2023 19:29:22 +0100
From:   Daniel Vetter <daniel@...ll.ch>
To:     Jiasheng Jiang <jiasheng@...as.ac.cn>
Cc:     jani.nikula@...ux.intel.com, joonas.lahtinen@...ux.intel.com,
        rodrigo.vivi@...el.com, tvrtko.ursulin@...ux.intel.com,
        airlied@...il.com, daniel@...ll.ch, ville.syrjala@...ux.intel.com,
        manasi.d.navare@...el.com, stanislav.lisovskiy@...el.com,
        lucas.demarchi@...el.com, intel-gfx@...ts.freedesktop.org,
        dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] drm/i915: Add missing check and destroy for
 alloc_workqueue

On Fri, Jan 06, 2023 at 05:09:34PM +0800, Jiasheng Jiang wrote:
> Add checks for the return value of alloc_workqueue and
> alloc_ordered_workqueue as they may return NULL pointer and cause NULL
> pointer dereference.
> Moreover, destroy them when fails later in order to avoid memory leak.
> Because in `drivers/gpu/drm/i915/i915_driver.c`, if
> intel_modeset_init_noirq fails, its workqueues will not be destroyed.
> 
> Fixes: c26a058680dc ("drm/i915: Use a high priority wq for nonblocking plane updates")
> Fixes: 757fffcfdffb ("drm/i915: Put all non-blocking modesets onto an ordered wq")
> Signed-off-by: Jiasheng Jiang <jiasheng@...as.ac.cn>

So you have an entire pile of these, and I think that's a really good
excuse to
- create a drmm_alloc_workqueue helper for these (and
  drmm_alloc_ordered_workqueue) using the drmm_add_action_or_reset()
  function for automatic drm_device cleanup
- use that instead in all drivers, which means you do not have to make any
  error handling mor complex

Up for that? In that case please also convert any existing alloc*workqueue
callsites in drm, and make this all a patch series (since then there would
be dependencies).

Cheers, Daniel


> ---
>  drivers/gpu/drm/i915/display/intel_display.c | 21 ++++++++++++++++----
>  1 file changed, 17 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> index 6c2686ecb62a..58f6840d6bd8 100644
> --- a/drivers/gpu/drm/i915/display/intel_display.c
> +++ b/drivers/gpu/drm/i915/display/intel_display.c
> @@ -8655,26 +8655,35 @@ int intel_modeset_init_noirq(struct drm_i915_private *i915)
>  	intel_dmc_ucode_init(i915);
>  
>  	i915->display.wq.modeset = alloc_ordered_workqueue("i915_modeset", 0);
> +	if (!i915->display.wq.modeset) {
> +		ret = -ENOMEM;
> +		goto cleanup_vga_client_pw_domain_dmc;
> +	}
> +
>  	i915->display.wq.flip = alloc_workqueue("i915_flip", WQ_HIGHPRI |
>  						WQ_UNBOUND, WQ_UNBOUND_MAX_ACTIVE);
> +	if (!i915->display.wq.flip) {
> +		ret = -ENOMEM;
> +		goto cleanup_modeset;
> +	}
>  
>  	intel_mode_config_init(i915);
>  
>  	ret = intel_cdclk_init(i915);
>  	if (ret)
> -		goto cleanup_vga_client_pw_domain_dmc;
> +		goto cleanup_flip;
>  
>  	ret = intel_color_init(i915);
>  	if (ret)
> -		goto cleanup_vga_client_pw_domain_dmc;
> +		goto cleanup_flip;
>  
>  	ret = intel_dbuf_init(i915);
>  	if (ret)
> -		goto cleanup_vga_client_pw_domain_dmc;
> +		goto cleanup_flip;
>  
>  	ret = intel_bw_init(i915);
>  	if (ret)
> -		goto cleanup_vga_client_pw_domain_dmc;
> +		goto cleanup_flip;
>  
>  	init_llist_head(&i915->display.atomic_helper.free_list);
>  	INIT_WORK(&i915->display.atomic_helper.free_work,
> @@ -8686,6 +8695,10 @@ int intel_modeset_init_noirq(struct drm_i915_private *i915)
>  
>  	return 0;
>  
> +cleanup_flip:
> +	destroy_workqueue(i915->display.wq.flip);
> +cleanup_modeset:
> +	destroy_workqueue(i915->display.wq.modeset);
>  cleanup_vga_client_pw_domain_dmc:
>  	intel_dmc_ucode_fini(i915);
>  	intel_power_domains_driver_remove(i915);
> -- 
> 2.25.1
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ