Laurent,
<Snip>
>I haven't tried to myself, it's just an idea that I'm throwing in the
>discussion.
One of our application engineer had done this and it seems to work fine. But he mentioned some issues with mmap (didn't remember what issue he had).
>
>> The driver is basically written to allocate multiple pools and buffers
>per
>> pool to satisfies various memory allocation requirements inside our TI
>SDK.
>
<Snip>
>Are you referring to the cmemk driver ?
Yes. CMEM.
It seems to really be a quick hack
>to
>short-circuit all the Linux kernel memory management infrastructure and let
>userspace allocate physical memory from a private pool of reserved SDRAM.
>I'd
>be quite surprised if something like that ever ends up in mainline.
>
Really. I looked at the code and it seems to do io_remap() on the entire memory region and do mmap of the buffer to user space. I am not sure why you think it is short circuiting the kernel memory management.
>Regards,
>
>Laurent Pinchart
>
淸{.n+돴윯돪+%듚깁負w{.n+돴{깸g돬왲^nr⊆z곷h솳鈺&｛z요z받쀺++zf＂h닱~넮녬i곴묎z_溫j:+v돣)山m
