[Intel-gfx] [PATCH] drm/i915/gem: Add a check for object size for corner cases

[Ram Moon, AnandX]

Hi Chris,

-----Original Message-----
From: dri-devel <dri-devel-bounces at lists.freedesktop.org> On Behalf Of Chris Wilson
Sent: Wednesday, February 10, 2021 4:15 PM
To: Ram Moon, AnandX <anandx.ram.moon at intel.com>; Jani Nikula <jani.nikula at linux.intel.com>; Auld, Matthew <matthew.auld at intel.com>; Surendrakumar Upadhyay, TejaskumarX <tejaskumarx.surendrakumar.upadhyay at intel.com>; Ursulin, Tvrtko <tvrtko.ursulin at intel.com>; dri-devel at lists.freedesktop.org; intel-gfx at lists.freedesktop.org
Cc: Ram Moon, AnandX <anandx.ram.moon at intel.com>
Subject: Re: [Intel-gfx] [PATCH] drm/i915/gem: Add a check for object size for corner cases

Quoting Anand Moon (2021-02-10 07:59:29)
> Add check for object size to return appropriate error -E2BIG or 
> -EINVAL to avoid WARM_ON and successful return for some testcase.

No. You miss the point of having those warnings. We need to inspect the code to remove the last remaining "int pagenum", and then we can remove the GEM_WARN_ON((sz) >> PAGE_SHIFT > INT_MAX). These are not emitted for users, only for us to motivate us into finally fixing the code.
-Chris

Yes, I got your point these check are meant to catch up integer overflow.

I have check with the IGT testcase case  _gem_create_ and _gem_userptr_blits_  
which fails pass size *-1ull << 32*  left shift and *0~* which leads to integer overflow 
ie  _negative_ size passed to create  i915_gem_create via ioctl  this leads to GM_WARN_ON. 

Can we drop these testcase so that we don't break the kernel ?

Thanks
-Anand
_______________________________________________
dri-devel mailing list
dri-devel at lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel