[PATCH] dma-pool: Do not allocate pool memory from CMA

Jeremy Linton jeremy.linton at arm.com
Wed Jul 8 19:14:39 EDT 2020


Hi,

On 7/8/20 11:49 AM, Nicolas Saenz Julienne wrote:
> There is no guarantee to CMA's placement, so allocating a zone specific
> atomic pool from CMA might return memory from a completely different
> memory zone. So stop using it.
> 
> Fixes: c84dc6e68a1d ("dma-pool: add additional coherent pools to map to gfp mask")
> Reported-by: Jeremy Linton <jeremy.linton at arm.com>
> Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne at suse.de>
> ---

With this patch and 
https://www.mail-archive.com/linux-kernel@vger.kernel.org/msg2226971.html 
the machine appears to be working fine with/without CMA and in both 
DT/ACPI mode.

The JBOD/etc I was having problems with are working fine, and the rtlsdr 
seems to be working better now too (its still not perfect, but that is 
likely another issue).

so:

tested-by: Jeremy Linton <jeremy.linton at arm.com>

thanks!


> 
> An more costly alternative would be adding an option to
> dma_alloc_from_contiguous() so it fails when the allocation doesn't fall
> in a specific zone.
> 
>   kernel/dma/pool.c | 11 ++---------
>   1 file changed, 2 insertions(+), 9 deletions(-)
> 
> diff --git a/kernel/dma/pool.c b/kernel/dma/pool.c
> index 8cfa01243ed2..4bc1c46ae6ef 100644
> --- a/kernel/dma/pool.c
> +++ b/kernel/dma/pool.c
> @@ -6,7 +6,6 @@
>   #include <linux/debugfs.h>
>   #include <linux/dma-direct.h>
>   #include <linux/dma-noncoherent.h>
> -#include <linux/dma-contiguous.h>
>   #include <linux/init.h>
>   #include <linux/genalloc.h>
>   #include <linux/set_memory.h>
> @@ -69,12 +68,7 @@ static int atomic_pool_expand(struct gen_pool *pool, size_t pool_size,
>   
>   	do {
>   		pool_size = 1 << (PAGE_SHIFT + order);
> -
> -		if (dev_get_cma_area(NULL))
> -			page = dma_alloc_from_contiguous(NULL, 1 << order,
> -							 order, false);
> -		else
> -			page = alloc_pages(gfp, order);
> +		page = alloc_pages(gfp, order);
>   	} while (!page && order-- > 0);
>   	if (!page)
>   		goto out;
> @@ -118,8 +112,7 @@ static int atomic_pool_expand(struct gen_pool *pool, size_t pool_size,
>   	dma_common_free_remap(addr, pool_size);
>   #endif
>   free_page: __maybe_unused
> -	if (!dma_release_from_contiguous(NULL, page, 1 << order))
> -		__free_pages(page, order);
> +	__free_pages(page, order);
>   out:
>   	return ret;
>   }
> 




More information about the linux-rpi-kernel mailing list