The armada_gem driver stores a DMA address token in a resource_size_t. This is safe because both are always at least 32 bits wide and can store any address that is accessible by this driver.
However, it is a slight type mismatch when we assign it from/to an sg_dma_address() value, and we get a warning in any configuration in which resource_size_t is shorter than dma_addr_t:
drivers/gpu/drm/armada/armada_gem.c: In function 'armada_gem_alloc_private_object': arch/arm/include/asm/dma-mapping.h:16:24: warning: large integer implicitly truncated to unsigned type [-Woverflow] #define DMA_ERROR_CODE (~(dma_addr_t)0x0) ^ drivers/gpu/drm/armada/armada_gem.c:204:18: note: in expansion of macro 'DMA_ERROR_CODE' obj->dev_addr = DMA_ERROR_CODE; ^
Changing the type to dma_addr_t solves both issues.
Signed-off-by: Arnd Bergmann arnd@arndb.de --- drivers/gpu/drm/armada/armada_gem.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/armada/armada_gem.h b/drivers/gpu/drm/armada/armada_gem.h index b000ea3a829a..270cff02c1cd 100644 --- a/drivers/gpu/drm/armada/armada_gem.h +++ b/drivers/gpu/drm/armada/armada_gem.h @@ -15,7 +15,7 @@ struct armada_gem_object { struct drm_gem_object obj; void *addr; phys_addr_t phys_addr; - resource_size_t dev_addr; + dma_addr_t dev_addr; struct drm_mm_node *linear; /* for linear backed */ struct page *page; /* for page backed */ struct sg_table *sgt; /* for imported */