Hi,
Lately i have been thinking to the memory allocation problem and i
believe that we should fix a maximum limit for BO size we allocate.
This limit should be decided by the kernel (i am assuming to be in a
KMS world).
Here is a small list of argument in favor of such idea :
- GPU have limited address space often fixed once the chipset
is initialized (for instance its GTT or VRAM size or VRAM+GTT)
- object GPU manipulate often have boundary for instance :
max texture size + max number of mipmap, or max 3d texture
size (as 3d texture is likely the biggest object i can think
of)
Anyone got idea against such limitations ?
Note that if i want upper limit is to be able to chunk down VRAM into
several bins of same size (bin size would be the upper limit for an
object). These can also be apply to GART. Also this upper limit could
be useful for determining maximum size a cmd buffer can reference.
Cheers,
Jerome Glisse
------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H
--
_______________________________________________
Dri-devel mailing list
Dri-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dri-devel
