Re: drivers/net/ethernet/toshiba/ps3_gelic_net.c:330:33: sparse: sparse: incorrect type in assignment (different base types)

[Geoff Levand]

Hi,

On 12/14/23 23:51, kernel test robot wrote:
> tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> head:   5bd7ef53ffe5ca580e93e74eb8c81ed191ddc4bd
> commit: bebe933d35a63d4f042fbf4dce4f22e689ba0fcd net/ps3_gelic_net: Use dma_mapping_error
> date:   9 months ago
> config: powerpc64-randconfig-r122-20231213 (https://download.01.org/0day-ci/archive/20231214/202312142250.7GMllD1E-lkp@xxxxxxxxx/config)
> compiler: clang version 17.0.0 (https://github.com/llvm/llvm-project.git 4a5ac14ee968ff0ad5d2cc1ffa0299048db4c88a)
> reproduce: (https://download.01.org/0day-ci/archive/20231214/202312142250.7GMllD1E-lkp@xxxxxxxxx/reproduce)
> 
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@xxxxxxxxx>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202312142250.7GMllD1E-lkp@xxxxxxxxx/
> 
> sparse warnings: (new ones prefixed by >>)
>>> drivers/net/ethernet/toshiba/ps3_gelic_net.c:330:33: sparse: sparse: incorrect type in assignment (different base types) @@     expected unsigned long long [usertype] bus_addr @@     got restricted __be32 [usertype] @@
>    drivers/net/ethernet/toshiba/ps3_gelic_net.c:330:33: sparse:     expected unsigned long long [usertype] bus_addr
>    drivers/net/ethernet/toshiba/ps3_gelic_net.c:330:33: sparse:     got restricted __be32 [usertype]
> 
> vim +330 drivers/net/ethernet/toshiba/ps3_gelic_net.c
> 
>    295	
>    296	/**
>    297	 * gelic_card_init_chain - links descriptor chain
>    298	 * @card: card structure
>    299	 * @chain: address of chain
>    300	 * @start_descr: address of descriptor array
>    301	 * @no: number of descriptors
>    302	 *
>    303	 * we manage a circular list that mirrors the hardware structure,
>    304	 * except that the hardware uses bus addresses.
>    305	 *
>    306	 * returns 0 on success, <0 on failure
>    307	 */
>    308	static int gelic_card_init_chain(struct gelic_card *card,
>    309					 struct gelic_descr_chain *chain,
>    310					 struct gelic_descr *start_descr, int no)
>    311	{
>    312		int i;
>    313		struct gelic_descr *descr;
>    314	
>    315		descr = start_descr;
>    316		memset(descr, 0, sizeof(*descr) * no);
>    317	
>    318		/* set up the hardware pointers in each descriptor */
>    319		for (i = 0; i < no; i++, descr++) {
>    320			dma_addr_t cpu_addr;
>    321	
>    322			gelic_descr_set_status(descr, GELIC_DESCR_DMA_NOT_IN_USE);
>    323	
>    324			cpu_addr = dma_map_single(ctodev(card), descr,
>    325						  GELIC_DESCR_SIZE, DMA_BIDIRECTIONAL);
>    326	
>    327			if (dma_mapping_error(ctodev(card), cpu_addr))
>    328				goto iommu_error;
>    329	
>  > 330			descr->bus_addr = cpu_to_be32(cpu_addr);

The values of the gelic hardware descriptors are all be32.  I think what we
want is to force cpu_addr to be32 so the conversion to be32 won't give a
warning, so something like this:

  descr->bus_addr = cpu_to_be32((__force __be32)cpu_addr);

I'll prepare a patch and submit after some testing.

-Geoff