linux-kernel - Re: [RESEND PATCH 4/5] drm: Add a drm_drv

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]

Date:   Thu, 4 Nov 2021 13:11:35 +0100
From:   Javier Martinez Canillas <javierm@...hat.com>
To:     Jani Nikula <jani.nikula@...ux.intel.com>,
        Thomas Zimmermann <tzimmermann@...e.de>,
        linux-kernel@...r.kernel.org
Cc:     Pekka Paalanen <pekka.paalanen@...labora.com>,
        David Airlie <airlied@...ux.ie>,
        Daniel Vetter <daniel.vetter@...ll.ch>,
        Michel Dänzer <michel@...nzer.net>,
        dri-devel@...ts.freedesktop.org,
        Peter Robinson <pbrobinson@...il.com>,
        Neal Gompa <ngompa13@...il.com>
Subject: Re: [RESEND PATCH 4/5] drm: Add a drm_drv_enabled() helper function

Hello Jani,

On 11/4/21 12:10, Jani Nikula wrote:
> On Wed, 03 Nov 2021, Thomas Zimmermann <tzimmermann@...e.de> wrote:
>> Hi
>>
>> Am 03.11.21 um 14:41 schrieb Jani Nikula:
>>> On Wed, 03 Nov 2021, Javier Martinez Canillas <javierm@...hat.com> wrote:
>>>> DRM drivers can use this to determine whether they can be enabled or not.
>>>>
>>>> For now it's just a wrapper around drm_modeset_disabled() but having the
>>>> indirection level will allow to add other conditions besides "nomodeset".
>>>>
>>>> Suggested-by: Thomas Zimmermann <tzimmermann@...e.de>
>>>> Signed-off-by: Javier Martinez Canillas <javierm@...hat.com>
>>>
>>> Can't see i915 trivially using this due to the drm_driver
>>> parameter. Please let's not merge helpers like this without actual
>>> users.
>>
>> Can you explain?
>>
>> This would be called on the top of the PCI probe function. The parameter 
>> would allow to decide on a per-driver base (e.g., to ignore generic 
>> drivers).
> 
> Where and how exactly? This is why we need to see the patch using the
> function. A patch is worth a thousand words. ;)
>

Thomas suggested to squash patches #4 and #5 so I'll do that when posting v2.
 
> See current vgacon_text_force() usage in i915/i915_module.c. It happens
> way before anything related to pci or drm driver. Why bother with the
> complicated setup and teardown of stuff if you can bail out earlier?
>

Yes, most drivers use vgacon_text_force() in their module init callback.

The ones that do in their probe function are drivers that don't have a
module init/exit but just use the module_platform_driver() macro.

I won't modify that and will keep the bail in the same place that the
drivers already do. I hope to have time and post a new revision today.

> 
> BR,
> Jani.
> 
Best regards,
-- 
Javier Martinez Canillas
Linux Engineering
Red Hat