Am 16.07.2014 18:12, schrieb Alex Deucher:
> On Wed, Jul 16, 2014 at 5:41 AM, Christian K?nig
> <deathsimple at vodafone.de> wrote:
>> Am 16.07.2014 11:40, schrieb Michel D?nzer:
>>
>>> From: Michel D?nzer <michel.daenzer at amd.com>
>>>
>>> These clutter up dmesg during piglit runs. Userspace generally deals
>>> gracefully with this failure.
>>>
>>> Signed-off-by: Michel D?nzer <michel.daenzer at amd.com>
>>
>> Reviewed-by: Christian K?nig <christian.koenig at amd.com>
>>
>> Already wanted to suggest the same thing,
> Applied to my 3.17 tree. Is there any reason we can't remove or relax
> this limit check in general? At the very least I think we can remove
> the visible vram limit.
Already thought about that as well. Since we fixed the fallback to GTT
if a buffer won't fit into VRAM on CPU access I offhand don't see any
obstacle to relaxing this test.
We should just test if that fallback can handle to larges buffers
gracefully.
Christian.
>
> Alex
>
>> Christian.
>>
>>
>>> ---
>>> drivers/gpu/drm/radeon/radeon_gem.c | 4 ++--
>>> 1 file changed, 2 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/radeon/radeon_gem.c
>>> b/drivers/gpu/drm/radeon/radeon_gem.c
>>> index 07292aa..8584e43 100644
>>> --- a/drivers/gpu/drm/radeon/radeon_gem.c
>>> +++ b/drivers/gpu/drm/radeon/radeon_gem.c
>>> @@ -58,8 +58,8 @@ int radeon_gem_object_create(struct radeon_device *rdev,
>>> int size,
>>> /* maximun bo size is the minimun btw visible vram and gtt size */
>>> max_size = min(rdev->mc.visible_vram_size, rdev->mc.gtt_size);
>>> if (size > max_size) {
>>> - printk(KERN_WARNING "%s:%d alloc size %dMb bigger than
>>> %ldMb limit\n",
>>> - __func__, __LINE__, size >> 20, max_size >> 20);
>>> + DRM_DEBUG("Allocation size %dMb bigger than %ldMb
>>> limit\n",
>>> + size >> 20, max_size >> 20);
>>> return -ENOMEM;
>>> }
>>>
>>
>> _______________________________________________
>> dri-devel mailing list
>> dri-devel at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/dri-devel
