order of matching alloc_pages/free_pages call pairs. Are they always same?

order of matching alloc_pages/free_pages call pairs. Are they always same?

[Bulent Abali]

Is it reasonable to assume that matching
alloc_pages/free_pages pairs will always have the same order
as the 2nd argument?

For example
pg = alloc_pages( , aorder);   free_pages(pg, forder);
Is (aorder == forder) always true?

Or, are there any bizarro drivers etc which will intentionally
free partial amounts, that is (forder < aorder)?


--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/

Re: order of matching alloc_pages/free_pages call pairs. Are they always same?

[Joseph A. Knapka]

Bulent Abali wrote:
> 
> Is it reasonable to assume that matching
> alloc_pages/free_pages pairs will always have the same order
> as the 2nd argument?
> 
> For example
> pg = alloc_pages( , aorder);   free_pages(pg, forder);
> Is (aorder == forder) always true?
> 
> Or, are there any bizarro drivers etc which will intentionally
> free partial amounts, that is (forder < aorder)?

This would be a somewhat bizarre thing to do, but after thinking
about it a bit, I believe it would work fine - as long as you're
very careful to free blocks with appropriate order and alignment.

I'm not personally aware of any code that actually frees sub-blocks
of allocated blocks, but I expect that when I get around to looking
at the slab allocator (kmalloc(), kfree()) there will be code
in there that does so.

-- Joe


-- Joseph A. Knapka
"If I ever get reincarnated... let me make certain I don't come back
 as a paperclip." -- protagonist, H Murakami's "Hard-boiled Wonderland"
// Linux MM Documentation in progress:
// http://home.earthlink.net/~jknapka/linux-mm/vmoutline.html
* Evolution is an "unproven theory" in the same sense that gravity is. *
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/

Re: order of matching alloc_pages/free_pages call pairs. Are they always same?

[Timur Tabi]

** Reply to message from "Bulent Abali" <abali@us.ibm.com> on Sat, 26 May 2001
13:10:42 -0400


> Is it reasonable to assume that matching
> alloc_pages/free_pages pairs will always have the same order
> as the 2nd argument?
> 
> For example
> pg = alloc_pages( , aorder);   free_pages(pg, forder);
> Is (aorder == forder) always true?
> 
> Or, are there any bizarro drivers etc which will intentionally
> free partial amounts, that is (forder < aorder)?

Yes!!  My driver does exactly that!!  Please DO NOT do anything to break this
functionality.  Here's the code which does that:

void fragment_and_free(void *_region, unsigned current_order, unsigned
new_order)
{
    unsigned long current_size = (1 << current_order) << PAGE_SHIFT;
    unsigned long new_size = (1 << new_order) << PAGE_SHIFT;

    char *region = (char *) _region;
    char *p;

    printk("Subdividing block of order %u into blocks of order %u\n",
current_order, new_order);

    if (new_order <= current_order)
    {
        for (p = region; p < (region + current_size); p += new_size)
            free_pages((u32) p, new_order);
    }
}


-- 
Timur Tabi - ttabi@interactivesi.com
Interactive Silicon - http://www.interactivesi.com

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/