mm/dmapool.c: revert "make dma pool to use kmalloc_node"
authorChristian König <christian.koenig@amd.com>
Fri, 14 Jan 2022 22:06:54 +0000 (14:06 -0800)
committerLinus Torvalds <torvalds@linux-foundation.org>
Sat, 15 Jan 2022 14:30:28 +0000 (16:30 +0200)
This reverts commit 2618c60b8b5836 ("dma: make dma pool to use
kmalloc_node").

While working myself into the dmapool code I've found this little odd
kmalloc_node().

What basically happens here is that we allocate the housekeeping
structure on the numa node where the device is attached to.  Since the
device is never doing DMA to or from that memory this doesn't seem to
make sense at all.

So while this doesn't seem to cause much harm it's probably cleaner to
revert the change for consistency.

Link: https://lkml.kernel.org/r/20211221110724.97664-1-christian.koenig@amd.com
Signed-off-by: Christian König <christian.koenig@amd.com>
Cc: Yinghai Lu <yinghai.lu@sun.com>
Cc: Andi Kleen <ak@suse.de>
Cc: Christoph Lameter <clameter@sgi.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/dmapool.c

index 64b537b..a7eb5d0 100644 (file)
@@ -152,7 +152,7 @@ struct dma_pool *dma_pool_create(const char *name, struct device *dev,
        else if ((boundary < size) || (boundary & (boundary - 1)))
                return NULL;
 
-       retval = kmalloc_node(sizeof(*retval), GFP_KERNEL, dev_to_node(dev));
+       retval = kmalloc(sizeof(*retval), GFP_KERNEL);
        if (!retval)
                return retval;