On 11/12/18 11:32 AM, John Garry wrote: > On 12/11/2018 15:42, Tony Battersby wrote: >> dmapool originally tried to support pools without a device because >> dma_alloc_coherent() supports allocations without a device. But nobody >> ended up using dma pools without a device, so the current checks in >> dmapool.c for pool->dev == NULL are both insufficient and causing bloat. >> Remove them. >> > As an aside, is it right that dma_pool_create() does not actually reject > dev==NULL and would crash from a NULL-pointer dereference? > > Thanks, > John > When passed a NULL dev, dma_pool_create() will already crash with a NULL-pointer dereference before this patch series, because it checks for dev == NULL in some places but not others. Specifically, it will crash in one of these two places in dma_pool_create(): retval = kmalloc_node(sizeof(*retval), GFP_KERNEL, dev_to_node(dev)); -or- if (list_empty(&dev->dma_pools)) So removing the checks for dev == NULL will not make previously-working code to start crashing suddenly. And since passing dev == NULL would be an API misuse error and not a runtime error, I would rather not add a new check to reject it. Tony
