Re: [patch 1/2] agpgart - allow user-populated memory types.

["Thomas Hellström" <thomas@tungstengraphics.com>]

Arjan van de Ven wrote:

>>A short background:
>>The current code uses vmalloc only. The potential use of kmalloc was 
>>introduced
>>to save memory and cpu-speed.
>>All agp drivers expect to see a single memory chunk, so I'm not sure we 
>>want to have an array of pages. That may require rewriting a lot of code.
>>    
>>
>
>but if it's clearly the right thing.....
>How hard can it be? there are what.. 5 or 6 AGP drivers in the kernel?
>
>
>  
>
Hmm,
but we would still waste a lot of memory compared to kmalloc,
when the amount of memory needed is much less than one page, which tends 
to be
a very common case.

Unless we allow the first entry in the array to be the virtual adress to an
arbitrary-sized (max one page) kmalloc() area, the rest of the entries 
can be pointers
to pages allocated with __get_free_page().

This would almost introduce the same level of confusion as the original 
proposal,
and effectively we'd be doing virtual address translation in software 
for each access.

>>If it's acceptable I'd like to go for the vmalloc / kmalloc flag, or at 
>>worst keep the current vmalloc only but that's such a _huge_ memory 
>>waste for small buffers. The flag was the original idea, but 
>>unfortunately the agp_memory struct is part of the drm interface, and I 
>>wasn't sure we could add a variable to it.
>>    
>>
>
>I doubt this is part of the userspace interface so for sure we can
>change it to be right.
>
>
>  
>
/Thomas