Re: [PATCH] virtio_ring: use alloc_pages_node for NUMA-aware allocation

[kernel test robot <lkp@intel.com>]

[-- Attachment #1: Type: text/plain, Size: 3767 bytes --]

Hi Shile,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on linus/master]
[also build test ERROR on v5.8-rc5 next-20200716]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Shile-Zhang/virtio_ring-use-alloc_pages_node-for-NUMA-aware-allocation/20200717-173734
base:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 8882572675c1bb1cc544f4e229a11661f1fc52e4
config: alpha-allyesconfig (attached as .config)
compiler: alpha-linux-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=alpha 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All error/warnings (new ones prefixed by >>):

   drivers/virtio/virtio_ring.c: In function 'vring_alloc_queue':
>> drivers/virtio/virtio_ring.c:280:52: error: passing argument 1 of 'dev_to_node' from incompatible pointer type [-Werror=incompatible-pointer-types]
     280 |   struct page *page = alloc_pages_node(dev_to_node(&vdev->dev.parent),
         |                                                    ^~~~~~~~~~~~~~~~~
         |                                                    |
         |                                                    struct device **
   In file included from include/linux/virtio.h:9,
                    from drivers/virtio/virtio_ring.c:6:
   include/linux/device.h:668:46: note: expected 'struct device *' but argument is of type 'struct device **'
     668 | static inline int dev_to_node(struct device *dev)
         |                               ~~~~~~~~~~~~~~~^~~
>> drivers/virtio/virtio_ring.c:284:4: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
     284 |    phys_addr_t phys_addr = virt_to_phys(queue);
         |    ^~~~~~~~~~~
   cc1: some warnings being treated as errors

vim +/dev_to_node +280 drivers/virtio/virtio_ring.c

   271	
   272	static void *vring_alloc_queue(struct virtio_device *vdev, size_t size,
   273				      dma_addr_t *dma_handle, gfp_t flag)
   274	{
   275		if (vring_use_dma_api(vdev)) {
   276			return dma_alloc_coherent(vdev->dev.parent, size,
   277						  dma_handle, flag);
   278		} else {
   279			void *queue = NULL;
 > 280			struct page *page = alloc_pages_node(dev_to_node(&vdev->dev.parent),
   281							     flag, get_order(size));
   282			if (page) {
   283				queue = page_address(page);
 > 284				phys_addr_t phys_addr = virt_to_phys(queue);
   285				*dma_handle = (dma_addr_t)phys_addr;
   286	
   287				/*
   288				 * Sanity check: make sure we dind't truncate
   289				 * the address.  The only arches I can find that
   290				 * have 64-bit phys_addr_t but 32-bit dma_addr_t
   291				 * are certain non-highmem MIPS and x86
   292				 * configurations, but these configurations
   293				 * should never allocate physical pages above 32
   294				 * bits, so this is fine.  Just in case, throw a
   295				 * warning and abort if we end up with an
   296				 * unrepresentable address.
   297				 */
   298				if (WARN_ON_ONCE(*dma_handle != phys_addr)) {
   299					free_pages_exact(queue, PAGE_ALIGN(size));
   300					return NULL;
   301				}
   302			}
   303			return queue;
   304		}
   305	}
   306	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 65028 bytes --]