Re: can device drivers return non-ram via vm_ops->nopage?

[Chris Wedgwood]

On Sun, Mar 21, 2004 at 06:23:22AM +0000, Christoph Hellwig wrote:

> Not sure how you get to fetchop here, but that driver does map ram
> pages so it should take pagefaults and not use remap_page_range().

It's been a while since I looked at this....  the fetchop driver maps
AMO space which is excluded from the EFI memory map (and any SHub
aliases) and thus shouldn't be touching anything normally considered
RAM.

<pause>

Checking the source I see:

    if (remap_page_range(vm_start, __pa(maddr), PAGE_SIZE, vma->vm_page_prot)) {
            fetchop_free_pages(vma->vm_private_data);
            vfree(vdata);
            fetchop_update_stats(-1, -pages);
            return -EAGAIN;
    }

as part of the drivers 'mmap fop'.  The underlying page is actually
from region-6 so I'm pretty sure it's safe.  If you think it is doing
something weird please let me know.



  --cw
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/